Digital integral evaluation method and system driven by civilized behavior data

Abstract

The invention discloses a digital integral evaluation method and system driven by civilized behavior data, and relates to the technical field of data processing and evaluation. Preprocessing the data; labeling the data; establishing a civilized behavior recognition model; training a civilized behavior recognition model; establishing a digital integral evaluation model: establishing an evaluation model based on a multi-modal feature fusion algorithm; the identified result in the step S5 is used as a data source to be input into the evaluation model; and outputting an evaluation result: outputting the evaluation result in a digital integral form according to the established evaluation model. According to the method, the civilized behaviors of the users can be represented by specific scores under the condition that different data sources of different users are known very conveniently through calculation, and the problem that the civilized behaviors are difficult to process and quantify is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing and evaluation, and in particular to a digital integral evaluation method and system driven by civilized behavior data. Background Art

[0002] Civilized behavior is a concept that covers a wide range of fields. It embodies the core socialist values, conforms to socialist morality and public order and good customs, and can lead social trends and promote social civilization and progress. Individual civilized behavior can promote the healthy development of the spiritual civilization of the entire society; therefore, promoting civilized development is also an important work goal of governments at all levels;

[0003] Civilized behavior itself is relatively abstract, and how to define civilized behavior mainly depends on people's subjective consciousness. However, with the scientific development of big data, it has become possible to use machine learning to identify civilized behavior. Therefore, quantifying the identified behavior itself and expressing the behavior itself in the form of points will help promote the promotion and development of civilized behavior and build civilized cities. Summary of the invention

[0004] The present invention is proposed in view of the above-mentioned background problems.

[0005] Therefore, the problem to be solved by the present invention is how to express different types of civilized behaviors in a unified scale in the form of digital integrals.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a digital score evaluation method driven by civilized behavior data, comprising the following specific steps:

[0007] S1. Data collection: Collect multi-source data information of authorized users and establish encrypted data sets according to the data sources; and set independent identity identification codes for users;

[0008] S2, data preprocessing: preprocess the multi-source data collected in step S1 to improve the accuracy of the original data;

[0009] S3. Data annotation: annotate the collected image or video information; divide it into civilized behavior samples and uncivilized behavior samples; extract text from the collected text information and classify it according to the behavior type;

[0010] S4. Establish a civilized behavior recognition model: Establish a civilized behavior evaluation model based on a deep learning neural network algorithm;

[0011] S5. Training civilized behavior recognition model: Divide the data set into a test set and a validation set to train the uncivilized behavior recognition model;

[0012] S6, establishing a digital integral evaluation model: establishing an evaluation model based on a multimodal feature fusion algorithm; inputting the identified result in step S5 as a data source into the evaluation model;

[0013] S7. Output evaluation results: Output the evaluation results in the form of digital integrals according to the established evaluation model.

[0014] Preferably, in step S1, the multi-source data information includes personal behavior record data and personal integrity record data;

[0015] Personal behavior records include video data collected by surveillance cameras, text and video data from social media platforms, and driving behavior habits;

[0016] Among them, surveillance cameras collect user behavior information to identify whether there is uncivilized behavior;

[0017] Obtain text, image, and video information sent by users through social media platforms to identify whether there is uncivilized behavior;

[0018] Driving behavior habit data is collected from the traffic control department database and is used to determine whether there is uncivilized behavior based on the frequency of the user's driving violations;

[0019] Personal integrity record data includes the user's environmental protection behavior, courageous behavior, and volunteer service behavior;

[0020] Environmental protection behavior data, courageous behavior data, and volunteer service data are recorded and the corresponding levels are determined based on the certification documents uploaded by the user.

[0021] Preferably, the pre-processing step in step S2 includes:

[0022] Data cleaning: remove duplicate, invalid and abnormal data;

[0023] Data normalization: normalize data from different sources and formats;

[0024] Image processing: rotation, mirror symmetry; adjusting brightness, height, contrast and sharpness; adding Gaussian noise.

[0025] Preferably, in S4, Yolo algorithm is used as a framework to extract features from the labeled data;

[0026] In the process of feature extraction, the geometric information of the low-level network of the image is integrated with the semantic information of the high-level network to improve the detection and segmentation performance.

[0027] Preferably, in step S5, the extracted image features are trained, the data set is first divided into a training set and a validation set, the images are classified based on a convolutional neural network (CNN) architecture, and predictions are made based on civilized behavior;

[0028] The F1-score on the validation set is used as the main indicator. When this indicator reaches above 0.9, the model is considered to have been trained to a good enough state; the training is completed;

[0029] The data of a single user is identified and the identification result is expressed in vector form:

[0030] P f (A1, A2...A i , B1, B2...B i )

[0031] Among them, P f is the user data set; A i is the integral value of the i-th source item in the personal behavior record data; B i is the integral value of the i-th source item in the personal credit record data;

[0032]

[0033] M i is the total amount of the i-th source project; N i is the amount of uncivilized behavior data in the i-th source project;

[0034]

[0035] Q i is the set of certification documents in the i-th source project in the personal credit record data; T k is the score coefficient of the certification document; h is a constant with a positive integer ranging from 5 to 10.

[0036] Preferably, in step S6, weight fusion is performed on data of different categories at the feature layer, and different weights are assigned to data of each category; and the obtained integral value is solved for the data based on the different weights;

[0037] The integral value obtained is calculated according to the following formula:

[0038]

[0039] Where H(x) is the integral value of the x-th user; n is the number of data sources, K is the score constant; P i is the recognition result of the i-th data source in the user dataset; C i For P i The corresponding data weight coefficient.

[0040] Preferably, the weight coefficient C i Related to the data type and amount of data, specifically:

[0041] The weight of personal behavior record data C Ai :

[0042] C Ai =(A i *M i ) / (A1*M1+A2*M2+......+A n *M n )

[0043] Among them, A i is the integral value of the i-th source item in the personal behavior record data; M i is the total amount of the i-th source item, and n is the number of items in the personal behavior record data;

[0044] The weight of personal credit record data C Bi :

[0045] C Bi =S i / (S1+S2+......+S n )

[0046] Among them, S i is the number of certification documents in the i-th source project in the personal credit record data; weight coefficient C i The following constraints are met:

[0047] ΣC Ai +ΣC Bi =1

[0048] If C Bi If both are not 0, then:

[0049] ΣC Ai ≤ΣC Bi

[0050] 0.5≤ΣC Bi <0.6.

[0051] A digital score evaluation system driven by civilized behavior data, used to execute the steps of the above method, including a data acquisition module, a data processing module, a civilized behavior evaluation module, and a digital score evaluation module;

[0052] The data collection module is used to collect multi-source information of authorized users; the multi-source information includes video data collected by surveillance cameras, text and video data from social media platforms, driving behavior habits, environmental protection behavior, courageous behavior and volunteer service behavior; and establish an encrypted database for users to store the above information;

[0053] The data processing module is used to pre-process the collected information, remove duplicate, invalid and abnormal data, and normalize the multi-source data;

[0054] The civilized behavior evaluation module is used to evaluate the normalized multi-source data and identify the uncivilized behaviors therein; the identification results are expressed in the form of vectors;

[0055] The digital integral evaluation module uses the recognition result of the civilized behavior evaluation module as input and calculates the value of the digital integral through integral operation.

[0056] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned civilized behavior data-driven digital score evaluation method when executing the computer program.

[0057] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned civilized behavior data-driven digital score evaluation method.

[0058] The beneficial effects of the present invention are:

[0059] 1. The present invention collects multi-source data of users, first improves the accuracy of the data through data processing and then labels the data; and uses the above data to establish an identification model for uncivilized behavior, and the types of identification include text, image or video; then establishes a system evaluation model; the present invention can very conveniently know through calculation that different users have different data sources, and can use specific scores to represent the civilized behavior of users, which solves the problem that civilized behavior itself is difficult to handle and quantify; in addition, users can adjust autonomously within the preset parameter range as needed. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0061] Figure 1 The present invention is a flowchart of a digital score evaluation method driven by civilized behavior data in an embodiment of the present invention.

[0062] Figure 2 4 is a structural diagram of a digital points evaluation system driven by civilized behavior data in an embodiment of the present invention. DETAILED DESCRIPTION

[0063] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0064] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0065] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0066] Example 1

[0067] A digital scoring evaluation method driven by civilized behavior data includes the following specific steps:

[0068] S1, data collection: collect multi-source data information of authorized users, and establish encrypted data sets according to the data sources; and set independent identity identification codes for users; the multi-source data information in step S1 includes personal behavior record data and personal integrity record data;

[0069] Personal behavior records include video data collected by surveillance cameras, text and video data from social media platforms, and driving behavior habits;

[0070] Among them, surveillance cameras collect user behavior information to identify whether there is uncivilized behavior;

[0071] Obtain text, image, and video information sent by users through social media platforms to identify whether there is uncivilized behavior;

[0072] Driving behavior habit data is collected from the traffic control department database and is used to determine whether there is uncivilized behavior based on the frequency of the user's driving violations;

[0073] Personal integrity record data includes the user's environmental protection behavior, courageous behavior, and volunteer service behavior;

[0074] Environmental protection behavior data, courageous behavior data, and volunteer service data are recorded and the corresponding levels are determined based on the certification documents uploaded by the user.

[0075] S2, data preprocessing: preprocess the multi-source data collected in step S1 to improve the accuracy of the original data;

[0076] The preprocessing steps in step S2 include:

[0077] Data cleaning: remove duplicate, invalid and abnormal data;

[0078] Data normalization: normalize data from different sources and formats;

[0079] Image processing: rotation, mirror symmetry; adjusting brightness, height, contrast and sharpness; adding Gaussian noise.

[0080] First, deduplicate the same type of data collected from different data sources, and then remove duplicate, invalid and abnormal data to improve the accuracy of the data source and the accuracy of subsequent model training. For example, for video or image data, existing AI recognition models (common ones include recognition models trained with TensorFlow Object Detection API or OpenPose) can be used to identify the content in the video and compare it with the information recorded in the collected text data. If the content similarity exceeds 80%, it is determined to be data duplication. Video or image data is deleted first, and text data is retained to reduce the amount of data storage and the difficulty of subsequent model training.

[0081] During the data normalization process, video files are uniformly converted to .mp4 format; image files are uniformly converted to .jpg format; and text data is uniformly converted to .xml format;

[0082] For images or video files, you can also improve the efficiency of feature recognition by rotating, mirroring, adjusting brightness, height, contrast and sharpness, and adding Gaussian noise.

[0083] Specific:

[0084] Randomly rotate the image, for example, choose 0°, 90°, 180°, or 270°. The rotation step can improve the model's robustness to different orientations.

[0085] Mirror symmetry increases the diversity of the data set and prevents overfitting.

[0086] Flip horizontally: flip the image left to right; Flip vertically: flip the image upside down (as needed).

[0087] Adjust the image data to a normal perspective, unify the input size of the image, and use the common size of 224x224 pixels;

[0088] Adjust brightness, height, contrast, and sharpness; turn blurry images into recognizable ones, and adjust out-of-proportion images to normal proportions and sizes; adjust brightness by increasing or decreasing the value of each pixel. Set a range from -50 (dark) to +50 (bright).

[0089] Adjusts the contrast by multiplying it by a factor ranging from 1.0 (no change) to 3.0 (increased contrast).

[0090] Use a convolution kernel to enhance the edges of the image. Use a 3x3 sharpening kernel. Make sure the image is resized to a 2:3 ratio.

[0091] Add Gaussian noise;

[0092] Adding Gaussian noise to the image can simulate the noise interference that may occur during the shooting process, which can enhance the model's robustness to noise and improve the accuracy and stability of recognition.

[0093] S3. Data annotation: annotate the collected image or video information; divide it into civilized behavior samples and uncivilized behavior samples; extract text from the collected text information and classify it according to the behavior type;

[0094] The data annotation process can be done manually or using existing video and image annotation tools; and then manually verified. After annotation, uncivilized behaviors are marked and distinguished; civilized behavior samples are used to establish data sets; uncivilized behavior samples are used to establish data sets according to categories: for example, littering, spitting, urinating and defecating, swearing, etc.

[0095] S4. Establish a civilized behavior recognition model: Establish a civilized behavior evaluation model based on a deep learning neural network algorithm; use the Yolo algorithm as a framework to extract features from the labeled data;

[0096] In the process of feature extraction, the geometric information of the low-level network of the image is integrated with the semantic information of the high-level network to improve the detection and segmentation performance.

[0097] S5. Training a civilized behavior recognition model: Divide the data set into a test set and a validation set. Preferably, the data in the civilized behavior samples and the uncivilized behavior samples can be used for validation; this is used to train an uncivilized behavior recognition model;

[0098] In step S5, the extracted image features are trained by first dividing the data set into a training set and a validation set, classifying the images based on the convolutional neural network (CNN) architecture, and predicting based on civilized behavior;

[0099] Specifically, specify the learning rate lr0 to be 0.01, the learning rate decay lrf to be 0.01, the momentum momentum to be 0.96, the weight decay weight_decay to be 0.001, the warmup training cycle warmup_epochs to be 6.0, the batch size batch to be 16, and the input image size imgsz to be 320;

[0100] The F1-score on the validation set is used as the main indicator (F1-score is the harmonic mean of precision and recall). When this indicator reaches above 0.9, the model is considered to have been trained to a sufficiently good state; the training is completed; after the civilized behavior evaluation model is trained, it can be used directly to identify uncivilized behavior in the data source without the need to train the model separately.

[0101] The data of a single user is identified and the identification result is expressed in vector form:

[0102] P f (A1, A2...A i , B1, B2...B i )

[0103] Among them, P f is the user data set; A i is the integral value of the i-th source item in the personal behavior record data; B i is the integral value of the i-th source item in the personal credit record data;

[0104]

[0105] M i is the total amount of the i-th source project; N i is the amount of uncivilized behavior data in the i-th source project;

[0106]

[0107] Q i is the set of certification documents in the i-th source project in the personal credit record data; T k is the score coefficient of the certification document; h is a constant with a positive integer ranging from 5 to 10.

[0108] In this embodiment, A1 represents the integral value of video data collected by the surveillance camera, A2 represents the integral value of text and video data on the social media platform, and A3 represents the integral value of driving behavior habits; B1 represents the integral value of environmental protection behavior data, B2 represents the integral value of courageous acts data, and B3 represents the integral value of volunteer service data. The civilized behavior is evaluated by digital points in six dimensions.

[0109] S6, establishing a digital integral evaluation model: establishing an evaluation model based on a multimodal feature fusion algorithm; inputting the identified result in step S5 as a data source into the evaluation model;

[0110] The data of different categories are weighted and fused at the feature layer, and different weights are assigned to each category of data; the integral value is obtained based on the different weights;

[0111] The integral value obtained is calculated according to the following formula:

[0112]

[0113] Where H(x) is the integral value of the x-th user; n is the number of data sources, K is the score constant; P i is the recognition result of the i-th data source in the user dataset; C i For P i The corresponding data weight coefficient.

[0114] Weight coefficient C i Related to the data type and amount of data, specifically:

[0115] The weight of personal behavior record data C Ai :

[0116] C Ai =(A i *M i ) / (A1*M1+A2*M2+......+A n *M n )

[0117] Among them, A i is the integral value of the i-th source item in the personal behavior record data; M i is the total amount of the i-th source item, and n is the number of items in the personal behavior record data;

[0118] The weight of personal credit record data C Bi :

[0119] C Bi =S i / (S1+S2+......+S n )

[0120] Where Si is the number of certification documents in the i-th source project in the personal credit record data; the weight coefficient C i The following constraints are met:

[0121] ΣC Ai +ΣC Bi =1

[0122] If CBi If both are not 0, then:

[0123] ΣC Ai ≤ΣC Bi

[0124] 0.5≤ΣC Bi <0.6.

[0125] S7. Output evaluation results: Output the evaluation results in the form of digital integrals according to the established evaluation model.

[0126] Example 2

[0127] The present invention provides a digital score evaluation system driven by civilized behavior data, which is used to execute the steps of the method in Example 1, including a data acquisition module, a data processing module, a civilized behavior evaluation module, and a digital score evaluation module;

[0128] The data collection module is used to collect multi-source information of authorized users; the multi-source information includes video data collected by surveillance cameras, text and video data from social media platforms, driving behavior habits, environmental protection behavior, courageous behavior and volunteer service behavior; and establish an encrypted database for users to store the above information;

[0129] The data processing module is used to pre-process the collected information, remove duplicate, invalid and abnormal data, and normalize multi-source data;

[0130] The civilized behavior evaluation module is used to evaluate the normalized multi-source data and identify the uncivilized behaviors therein; the identification results are expressed in the form of vectors;

[0131] The digital integral evaluation module uses the recognition result of the civilized behavior evaluation module as input and calculates the value of the digital integral through integral operation.

[0132] The following is an introduction to the solution using a real case. The specific operations are as follows:

[0133] First, use the data collection module to collect multi-source data information of 5 authorized users, and replace the 5 users with letters A, B, C, D, and E in turn;

[0134] Table 1 shows the multi-source data collection of 5 users

[0135]

[0136]

[0137] The video data collected by the surveillance camera in the above data refers to the number of times the user passes through the video surveillance area per unit time, and the data volume of uncivilized behavior refers to the number of times the uncivilized behavior occurs;

[0138] The video media platform data includes the total amount of text, pictures, videos, replies, and comments posted by users. The amount of uncivilized behavior data includes the amount of uncivilized text, pictures, videos, and replies to comments.

[0139] Driving behavior refers to the number of days in the past year, and the number of uncivilized behaviors refers to the number of violations in a year.

[0140] Environmental protection behavior, courageous behavior and volunteer service behavior are based on the corresponding service certificates and grades officially issued by the government departments. k The score coefficient for the certification document is calculated as 5 points for level one, 2 points for level two, and 1 point for level three.

[0141] Substitute the above-obtained data into the formula to calculate the values ​​of Ai and Bi

[0142]

[0143] M i is the total amount of the i-th source project; N i is the amount of uncivilized behavior data in the i-th source project;

[0144]

[0145] Q i is the set of certification documents in the i-th source project in the personal credit record data; h is a constant with a positive integer ranging from 5 to 10, and here h is 10 for calculation.

[0146] The results are as follows:

[0147] P fA (0.92, 0.93, 0.99, 1.2, 0, 3.5), P fB (0.97, 0.94, 1, 0, 0, 1.6), P fC (0.94, 1, 0.98, 2.5, 0, 0), P fD (0.93, 0.99, 1, 0, 1.5, 1.1), P fE (1, 0.99, 1, 0, 0, 1.2).

[0148] Then calculate the weight coefficient according to the data in Table 1.

[0149] The weight of personal behavior record data C Ai :

[0150] C Ai =(A i *M i) / (A1*M1+A2*M2+.......+A n *M n )

[0151] Among them, A i is the integral value of the i-th source item in the personal behavior record data; M i is the total amount of the i-th source item, and n is the number of items in the personal behavior record data;

[0152] The weight of personal credit record data C Bi :

[0153] C Bi =S i / (S1+S2+......+S n )

[0154] Among them, S i is the number of certification documents in the i-th source project in the personal credit record data; weight coefficient C i The following constraints are met:

[0155] ΣC Ai +ΣC Bi =1

[0156] If C Bi If both are not 0, then:

[0157] ΣC Ai ≤ΣC Bi

[0158] 0.5≤ΣC Bi <0.6.

[0159] Then substitute the calculated weight coefficient into H(x) to calculate the integral value of each user;

[0160]

[0161] Where H(x) is the integral value of the x-th user; n is the number of data sources, K is the score constant, which is 80 here; P i is the recognition result of the i-th data source in the user dataset; C i For P i The corresponding data weight coefficient.

[0162] After calculation, the weight coefficient of user A is (0.02, 0.135, 0.34, 0.08, 0, 0.415), and H(A)=452.576

[0163] After calculation, the weight coefficient of user B is (0.03, 0.19, 0.27, 0, 0, 0.5) and H(B)=139.6

[0164] After calculation, the weight coefficient of user C is (0.045, 0.0675, 0.355, 0.5, 0, 0) and H(C)=285.72

[0165] After calculation, the weight coefficient of user D is (0.013, 0.164, 0.322, 0, 0.25, 0.25) and H(D)=116.48

[0166] After calculation, the weight coefficient of user E is (0.015, 0.19, 0.29, 0, 0, 0.5), and H(E)=96.4.

[0167] In this embodiment, it is very convenient to know through calculation that different users have different data sources, and the civilized behavior of users can be represented by specific scores, which solves the problem that civilized behavior itself is difficult to handle and quantify; in addition, users can make independent adjustments within the preset parameter range as needed.

[0168] Example 3

[0169] This embodiment also provides a computer device, which is suitable for the digital score evaluation method driven by civilized behavior data, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the digital score evaluation method driven by civilized behavior data as proposed in the above embodiment.

[0170] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0171] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the multi-dimensional positioning and safety detection method based on a safety helmet as proposed in the above embodiment is implemented.

[0172] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A digital scoring evaluation method driven by civilized behavior data, characterized by: The specific steps include: S1. Data collection: Collect multi-source data information of authorized users and establish encrypted data sets according to the data sources; and set independent identity identification codes for users; S2, data preprocessing: preprocess the multi-source data collected in step S1 to improve the accuracy of the original data; S3. Data annotation: annotate the collected image or video information; divide it into civilized behavior samples and uncivilized behavior samples; extract text from the collected text information and classify it according to the behavior type; S4. Establish a civilized behavior recognition model: Establish a civilized behavior evaluation model based on a deep learning neural network algorithm; S5. Training civilized behavior recognition model: Divide the data set into a test set and a validation set to train the uncivilized behavior recognition model; S6, establishing a digital integral evaluation model: establishing an evaluation model based on a multimodal feature fusion algorithm; inputting the identified result in step S5 as a data source into the evaluation model; S7. Output evaluation results: Output the evaluation results in the form of digital integrals according to the established evaluation model.

2. The digital scoring evaluation method driven by civilized behavior data as claimed in claim 1, characterized in that: In step S1, the multi-source data information includes personal behavior record data and personal integrity record data; Personal behavior records include video data collected by surveillance cameras, text and video data from social media platforms, and driving behavior habits; Among them, surveillance cameras collect user behavior information to identify whether there is uncivilized behavior; Obtain text, image, and video information sent by users through social media platforms to identify whether there is uncivilized behavior; Driving behavior habit data is collected from the traffic control department database and is used to determine whether there is uncivilized behavior based on the frequency of the user's driving violations; Personal integrity record data includes the user's environmental protection behavior, courageous behavior, and volunteer service behavior; Environmental protection behavior data, courageous behavior data, and volunteer service data are recorded and the corresponding levels are determined based on the certification documents uploaded by the user.

3. The digital scoring evaluation method driven by civilized behavior data as claimed in claim 1, characterized in that: The preprocessing steps in step S2 include: Data cleaning: remove duplicate, invalid and abnormal data; Data normalization: normalize data from different sources and formats; Image processing: rotation, mirror symmetry; adjusting brightness, height, contrast and sharpness; adding Gaussian noise.

4. The digital scoring evaluation method driven by civilized behavior data as claimed in claim 1, characterized in that: In S4, the Yolo algorithm is used as a framework to extract features from the labeled data; In the process of feature extraction, the geometric information of the low-level network of the image is integrated with the semantic information of the high-level network to improve the detection and segmentation performance.

5. The digital scoring evaluation method driven by civilized behavior data as claimed in claim 1, characterized in that: In step S5, the extracted image features are trained by first dividing the data set into a training set and a validation set, classifying the images based on the convolutional neural network (CNN) architecture, and predicting based on civilized behavior; The F1-score on the validation set is used as the main indicator. When this indicator reaches above 0.9, the model is considered to have been trained to a good enough state; the training is completed; The data of a single user is identified and the identification result is expressed in vector form: <h2 style=";text-align:left;direction:ltr">P<h2 style=";text-align:left;direction:ltr"> f <h2 style=";text-align:left;direction:ltr"> (A1, A2……A<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> B1, B2……B<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> ) Among them, P f is the user data set; A i is the integral value of the i-th source item in the personal behavior record data; B i is the integral value of the i-th source item in the personal credit record data; M i is the total amount of the i-th source project; N i is the amount of uncivilized behavior data in the i-th source project; Q i is the set of certification documents in the i-th source project in the personal credit record data; T k is the score coefficient of the certification document; h is a constant with a positive integer ranging from 5 to 10.

6. The digital scoring evaluation method driven by civilized behavior data as claimed in claim 1, characterized in that: In step S6, different categories of data are weighted and fused at the feature layer, and different weights are assigned to each category of data; the data are solved based on the different weights to obtain the integral value; The integral value obtained is calculated according to the following formula: Where H(x) is the integral value of the x-th user; n is the number of data sources, K is the score constant; P i is the recognition result of the i-th data source in the user dataset; C i For P i The corresponding data weight coefficient.

7. The digital scoring evaluation method driven by civilized behavior data as claimed in claim 6, characterized in that: Weight coefficient C i Related to data type and data amount: The weight of personal behavior record data C Ai : C Ai =(A i *M i ) / (A1*M1+A2*M2+......+A n *M n ) Among them, A i is the integral value of the i-th source item in the personal behavior record data; M i is the total amount of the i-th source item, and n is the number of items in the personal behavior record data; The weight of personal credit record data C Bi : C Bi =S i / (S1+S2+......+S n ) Among them, S i is the number of certification documents in the i-th source project in the personal credit record data; weight coefficient C i The following constraints are met: ∑C Ai +∑C Bi =1 If C Bi If both are not 0, then: ∑C Ai ≤∑C Bi 0.5≤∑C Bi ＜0.6。 8. A digital scoring evaluation system driven by civilized behavior data, used to perform the steps of any one of the methods of claims 1 to 7, characterized in that: It includes data collection module, data processing module, civilized behavior evaluation module and digital points evaluation module; The data collection module is used to collect multi-source information of authorized users; the multi-source information includes video data collected by surveillance cameras, text and video data of social media platforms, driving behavior habits, environmental protection behavior, courageous behavior and volunteer service behavior; And establish an encrypted database for users to store the above information; The data processing module is used to pre-process the collected information, remove duplicate, invalid and abnormal data, and normalize the multi-source data; The civilized behavior evaluation module is used to evaluate the normalized multi-source data and identify the uncivilized behaviors therein; the identification results are expressed in the form of vectors; The digital integral evaluation module uses the recognition result of the civilized behavior evaluation module as input and calculates the value of the digital integral through integral operation.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the digital score evaluation method driven by civilized behavior data according to any one of claims 1 to 7 are implemented.

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 digital scoring evaluation method driven by civilized behavior data as described in any one of claims 1 to 7 is implemented.

Images