Risk portrait assessment and risk decision system and computer equipment

Abstract

The invention relates to a risk portrait assessment and risk decision system and computer equipment. The data preprocessing module is used for preprocessing the initial image data, the initial text data and the initial sensing data to obtain target image data, target text data and target sensing data; the risk portrait evaluation module adopts an image risk portrait evaluation model to process the target image data to obtain an image risk portrait, and adopts a text risk portrait evaluation model to process the target text data to obtain a text risk portrait; preprocessing the target sensing data by using a sensing risk portrait evaluation model to obtain a sensing risk portrait, and fusing the obtained risk portraits to obtain a target risk portrait; and the risk decision module determines a coping strategy of the target risk portrait and future potential risks according to the target risk portrait, the multi-source data and the historical multi-source data. By adopting the method, the accuracy of the determined coping strategy of the security problem and the future potential risk can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of data analysis, and particularly to a risk portrait evaluation and risk decision-making system and a computer device. Background Art

[0002] In today's society, with the rapid development of technology and the acceleration of the urbanization process, people's living and working environments have become increasingly complex. This complexity not only brings convenience but also various potential safety risks. Especially in public places, industrial production, transportation and other fields, safety issues are particularly prominent.

[0003] In related technologies, cameras are often arranged in some places to shoot the places through the cameras. Business personnel discover safety problems at the scene through the images captured by the cameras, and analyze the coping strategies for the safety problems and potential future risks based on the captured images according to experience.

[0004] However, in related technologies, there are situations of missed judgments and misjudgments for safety problems, and the coping strategies for the safety problems and potential future risks are often analyzed based on experience, resulting in the inability to accurately determine the coping strategies for safety problems and potential future risks. Summary of the Invention

[0005] Based on this, the present application provides a risk portrait evaluation and risk decision-making system and a computer device, which can improve the accuracy of determining the coping strategies for safety problems and potential future risks.

[0006] In a first aspect, the present application provides a risk portrait evaluation and risk decision-making system, and the system includes:

[0007] A data acquisition module, configured to obtain multi-source data from a database;

[0008] A data preprocessing module, configured to obtain initial image data, initial text data, and initial sensing data from the multi-source data, preprocess the initial image data to obtain target image data, preprocess the initial text data to obtain target text data, and preprocess the initial sensing data to obtain target sensing data;

[0009] A risk portrait evaluation module, configured to process the target image data by using an image risk portrait evaluation model to obtain an image risk portrait, process the target text data by using a text risk portrait evaluation model to obtain a text risk portrait, preprocess the target sensing data by using a sensing risk portrait evaluation model to obtain a sensing risk portrait, and fuse the image risk portrait, the text risk portrait, and the sensing risk portrait to obtain a target risk portrait;

[0010] A risk decision-making module, configured to determine a coping strategy for the target risk profile and potential future risks according to the target risk profile, the multi-source data, and historical multi-source data.

[0011] In some embodiments, the system further includes:

[0012] A question-and-answer module, configured to obtain a user query request, where the user query request carries query location information and a query event. When the query event is included in the target risk profile, output the coping strategy for the query event in the target risk profile; when the query event is not included in the target risk profile, output an indication information indicating that the query event does not exist at the query location.

[0013] In some embodiments, the system further includes:

[0014] A voice broadcast module, configured to determine risk broadcast information corresponding to the target risk profile, obtain the type of a computer device having the system, determine at least one language type for broadcast according to the type of the computer device, convert the risk broadcast information using the languages of the respective language types to obtain respective target play information, and perform voice broadcast on the respective target play information.

[0015] In some embodiments, the data preprocessing module includes:

[0016] A data cleaning sub-module, configured to perform image duplicate removal processing and image denoising processing on the initial image data to obtain preset image data, perform text cleaning processing on the initial text data to obtain preset text data, and perform data cleaning processing on the initial sensing data to obtain preset sensing data;

[0017] A data format conversion sub-module, configured to perform image format conversion on the preset image data to obtain set image data in a set image format, perform text format conversion on the preset text data to obtain set text data in a set text format, and perform data format conversion on the preset sensing data to obtain set sensing data in a set data format;

[0018] A data standardization sub-module, configured to perform image size adjustment processing on the set image data to obtain target image data with a preset image size, perform standardization processing on the set text data to obtain target text data, and perform standardization processing on the set sensing data to obtain target sensing data with a preset dimension and a preset range.

[0019] In some embodiments, the risk profile evaluation module includes:

[0020] A risk profile assessment sub-module for processing the target image data using an image risk profile assessment model to obtain an image risk profile, processing the target text data using a text risk profile assessment model to obtain a text risk profile, and preprocessing the target sensing data using a sensing risk profile assessment model to obtain a sensing risk profile;

[0021] A risk profile fusion sub-module for fusing the image risk profile, the text risk profile, and the sensing risk profile to obtain a target risk profile.

[0022] In some embodiments, the risk profile assessment sub-module includes:

[0023] An image risk profile assessment unit for determining at least two image risk assessment dimensions, using respective image feature extraction modules corresponding to the image risk assessment dimensions to separately extract features from the target image data to obtain respective image feature information, processing the respective image feature information using an image feature assessment model to obtain assessment results under the respective image risk assessment dimensions, and determining the image risk profile according to the assessment results under the respective image risk assessment dimensions;

[0024] A text risk profile assessment unit for determining at least two text risk assessment dimensions, using respective text feature extraction modules corresponding to the text risk assessment dimensions to separately extract features from the target text data to obtain respective text feature information, processing the respective text feature information using a text feature assessment model to obtain assessment results under the respective text risk assessment dimensions, and determining the text risk profile according to the assessment results under the respective text risk assessment dimensions;

[0025] A sensing risk profile assessment unit for determining at least two sensing risk assessment dimensions, using respective sensing feature extraction modules corresponding to the sensing risk assessment dimensions to separately extract features from the target sensing data to obtain respective sensing feature information, processing the respective sensing feature information using a sensing feature assessment model to obtain assessment results under the respective sensing risk assessment dimensions, and determining the sensing risk profile according to the assessment results under the respective sensing risk assessment dimensions.

[0026] In some embodiments, the risk profile fusion sub-module includes:

[0027] An assessment result fusion unit for fusing the assessment results under the respective image risk assessment dimensions, the assessment results under the respective text risk assessment dimensions, and the assessment results under the respective sensing risk assessment dimensions to obtain assessment results under multiple comprehensive risk assessment dimensions;

[0028] A risk profile fusion unit is configured to determine the weights under each comprehensive risk assessment dimension according to the weights under each of the obtained image risk assessment dimensions, the weights under each of the text risk assessment dimensions, and the weights under each of the sensing risk assessment dimensions, and perform weighted fusion on the evaluation results under each comprehensive risk assessment dimension and the weights under each comprehensive risk assessment dimension to obtain the target risk profile.

[0029] In some embodiments, the risk decision module includes:

[0030] A rule inference sub-module is configured to input the target risk profile and the multi-source data into a rule inference model, and output a coping strategy for the target risk profile through the rule inference model;

[0031] A probability inference sub-module is configured to input the multi-source data, historical multi-source data, and the target risk profile into a probability inference model, and output potential future risks through the probability inference model.

[0032] In some embodiments, the system further includes a data prediction module: The data prediction module is configured to use an image prediction model to predict the initial image data to obtain predicted image data, use a text prediction model to predict the initial text data to obtain predicted text data, and use a sensing prediction model to predict the initial sensing data to obtain predicted sensing data;

[0033] A data preprocessing module is configured to preprocess the predicted image data and the initial image data to obtain the target image data, preprocess the predicted text data and the initial text data to obtain the target text data, and preprocess the predicted sensing data and the initial sensing data to obtain the target sensing data.

[0034] In a second aspect, the present application provides a computer device, which includes the risk profile evaluation and risk decision system according to any one of the above. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required to be used in the description of the embodiments of the present application or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0036] Figure 1 FIG. is a schematic structural diagram of a risk profile evaluation and risk decision system provided in the first embodiment;

[0037] Figure 2 Structural schematic diagram of a risk portrait evaluation and risk decision-making system provided for the second embodiment;

[0038] Figure 3 Structural schematic diagram of a risk portrait evaluation and risk decision-making system provided for the third embodiment;

[0039] Figure 4 Structural schematic diagram of a risk portrait evaluation and risk decision-making system provided for the fourth embodiment;

[0040] Figure 5 Structural schematic diagram of a risk portrait evaluation and risk decision-making system provided for the fifth embodiment;

[0041] Figure 6 Structural schematic diagram of a risk portrait evaluation and risk decision-making system provided for the sixth embodiment;

[0042] Figure 7 Structural schematic diagram of a risk portrait evaluation and risk decision-making system provided for the seventh embodiment;

[0043] Figure 8 Structural schematic diagram of a risk portrait evaluation and risk decision-making system provided for the eighth embodiment;

[0044] Figure 9 Structural schematic diagram of a multi-dimensional information fusion personal safety portrait and intelligent decision-making system provided for some embodiments;

[0045] Figure 10 Structural schematic diagram of a computer device provided for some embodiments. Detailed implementation manners

[0046] Hereinafter, embodiments of the technical solutions of the present application will be described in detail with reference to the accompanying drawings. The following embodiments are only used to illustrate the technical solutions of the present application more clearly, and therefore are only examples and cannot be used to limit the protection scope of the present application.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above accompanying drawing descriptions are intended to cover non-exclusive inclusion.

[0048] In the description of the embodiments of the present application, technical terms such as "first" and "second" are only used to distinguish different objects, and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity, specific order or primary-secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "a plurality of" is more than two, unless otherwise specifically defined. In the description of the embodiments of the present application, "each" means each one or each one among a plurality, unless otherwise specifically defined.

[0049] Reference to "embodiment" herein means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0050] In the description of the embodiments of the present application, the term "and / or" is merely a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0051] With the development of Internet of Things technology, sensor technology has been widely applied in the field of security monitoring. Sensors can be deployed in various environments to monitor parameters such as temperature, humidity, smoke, and vibration in real time, providing data support in more dimensions for security monitoring. However, the data of a single sensor often cannot comprehensively reflect the security situation and needs to be fused and analyzed with other data sources to more accurately identify risks.

[0052] As an important platform for information dissemination, social media data contains a large amount of clues and information about public security. By analyzing the content such as remarks, pictures, and videos on social media, the concerns and potential risks of the public regarding security issues can be discovered in a timely manner. However, the amount of social media data is huge and chaotic, and effective extraction and analysis need to be carried out with the help of natural language processing technology.

[0053] Currently, there are some security monitoring and early warning systems on the market, but most of these systems are limited to a single data source or a single function, lacking the ability to comprehensively analyze and process multi-dimensional information. For example, some systems only rely on video surveillance data for security monitoring, while ignoring other important data sources such as sensor data and social media data. In addition, the systems in related technologies often lack an intelligent decision support function and cannot provide effective coping strategies and suggestions for managers.

[0054] The computer devices in the embodiments of the present application may include one or a combination of at least two of the following: servers, mobile phones, tablets (Pads), computers with transceiver functions, handheld computers, desktop computers, personal digital assistants, portable media players, smart speakers, navigation devices, smart watches, smart glasses, wearable devices such as smart necklaces, pedometers, digital TVs, virtual reality (VR) devices, augmented reality (AR) devices, devices in industrial control, devices in self-driving, devices in remote medical surgery, devices in smart grid, devices in transportation safety, devices in smart city, devices in smart home, vehicles, in-vehicle devices, in-vehicle modules, and so on.

[0055] Figure 1 A structural schematic diagram of a risk portrait evaluation and risk decision-making system provided for the first embodiment. The risk portrait evaluation and risk decision-making system includes a data acquisition module, a data preprocessing module, a risk portrait evaluation module, and a risk decision-making module. Among them, the data acquisition module is connected to the data preprocessing module, the data preprocessing module is further connected to the risk portrait evaluation module, and the risk portrait evaluation module is further connected to the risk decision-making module.

[0056] The data acquisition module is used to obtain multi-source data from the database.

[0057] The data preprocessing module is used to obtain initial image data, initial text data, and initial sensing data from the multi-source data, preprocess the initial image data to obtain target image data, preprocess the initial text data to obtain target text data, and preprocess the initial sensing data to obtain target sensing data.

[0058] The risk portrait evaluation module is used to process the target image data using an image risk portrait evaluation model to obtain an image risk portrait, process the target text data using a text risk portrait evaluation model to obtain a text risk portrait, preprocess the target sensing data using a sensing risk portrait evaluation model to obtain a sensing risk portrait, and fuse the image risk portrait, the text risk portrait, and the sensing risk portrait to obtain a target risk portrait.

[0059] The risk decision-making module is used to determine the coping strategy for the target risk profile and potential future risks based on the target risk profile, the multi-source data, and historical multi-source data.

[0060] Multi-source data can be data obtained from multiple sources. Exemplarily, multi-source data can include at least one of the following: image data collected by a camera, image data converted from video data collected by a camera, image data in media, text data in media, image data converted from video data in media, text data converted from media audio data, image data uploaded by a user's personal device, text data uploaded by a user's personal device, sensing data uploaded by a user's personal device, and data collected by sensors arranged at the target site. That is, multi-source data includes image data, text data, and sensing data.

[0061] For example, the user's personal device can be at least one of a mobile phone, a smartwatch, a computer, etc. For example, the sensors can include a temperature sensor, a humidity sensor, a smoke sensor, a pressure sensor, a vibration sensor, a physiological sensor, etc. For example, the physiological sensor can include at least one of a heart rate sensor, a blood pressure sensor, a blood oxygen saturation sensor, a blood glucose sensor, etc. The multi-source data can be stored in a database. In some embodiments, the multi-source data can be real-time data.

[0062] The data acquisition module can be communicatively connected to the database. The data acquisition module can obtain multi-source data from the database. Exemplarily, the data acquisition module can obtain the multi-source data generated within a preset time period from the database every preset time period. Additionally, the data acquisition module can obtain the multi-source data generated within a preset time period from the database in response to a decision instruction initiated by the user. For example, the preset time period can be 1 minute, 5 minutes, 10 minutes, etc.

[0063] Exemplarily, the multi-source data including image data, text data, and sensing data can be respectively determined as initial image data, initial text data, and initial sensing data.

[0064] Exemplarily, data preprocessing can include at least one of the following processes: data cleaning process, data format conversion process, data standardization process.

[0065] A risk profile is the result of multi-dimensional description and analysis of the risk characteristics of a certain risk subject or event. The risk profile can be a description of personal risks. The risk profile can include the scores of each risk assessment dimension in multiple risk assessment dimensions and the score of the comprehensive risk assessment.

[0066] For example, the risk profile can be a risk profile for an individual participating in a large-scale sports event. The risk profile can include scores for the venue safety dimension, the event type dimension, and the individual location dimension, as well as a comprehensive personal risk score determined based on these three scores. The comprehensive personal risk score represents the probability of an individual facing personal risks. The risk profile can be a description of the venue risks. For example, the risk profile assessment and risk decision-making system is applied to a computer device, so that the risk profile of an individual using the computer device can be obtained through this system.

[0067] Also for example, the risk profile can be a risk profile for enterprise network security. The profile can include scores for the system vulnerability dimension, the data security dimension, and the network protection dimension, as well as a comprehensive enterprise network risk score determined based on these three scores. The comprehensive enterprise network risk score represents the probability of the enterprise network facing network failure risks. For example, the risk profile assessment and risk decision-making system is applied to an enterprise, so that the risk profile of the enterprise can be obtained through this system.

[0068] In some embodiments, when the probability of facing risks represented by the target risk profile is greater than or equal to a preset probability, the response strategy for the target risk profile and future potential risks can be determined based on the target risk profile, the multi-source data, and the historical multi-source data; when the probability is less than the preset probability, the future potential risks can be determined based on the target risk profile, the multi-source data, and the historical multi-source data.

[0069] Exemplarily, the response strategy for the target risk profile can be determined based on the target risk profile and the multi-source data, and the future potential risks can be determined based on the target risk profile, the multi-source data, and the historical multi-source data.

[0070] In some embodiments, the risk profile assessment and risk decision-making system can include a display module. The display module can display at least one of the following: the target risk profile, the response strategy for the target risk profile, and future potential risks. The display module can be connected to the risk profile assessment module and can also be connected to the risk decision module.

[0071] In the embodiments of the present application, a data preprocessing module preprocesses various types of data to obtain preprocessed data. A risk profile evaluation module processes the preprocessed data using a risk profile evaluation model to obtain a target risk profile. A risk profile is the result of multi-dimensional description and analysis of the risk characteristics of a certain risk subject or event, thereby avoiding the situation where business personnel miss or misjudge security problems due to fatigue and distraction, resulting in inaccurate discovery of security problems. Therefore, the response strategy for security problems and future potential risks can be accurately determined through the identified target risk profile, thereby improving the accuracy of determining the response strategy for security problems and future potential risks. The risk decision-making module determines the response strategy for the target risk profile and future potential risks based on the target risk profile, the multi-source data, and historical multi-source data, avoiding the inaccurate analysis caused by using empirical analysis for the response strategy for security problems and future potential risks, thereby further improving the accuracy of determining the response strategy for security problems and future potential risks. Moreover, since the data acquisition module obtains multi-source data, the data source is more comprehensive, thereby improving the accuracy of the determined target risk profile through comprehensive data, and thus improving the accuracy of determining the response strategy for security problems and future potential risks.

[0072] Figure 2 FIG. 4 is a schematic structural diagram of a risk profile evaluation and risk decision-making system provided in the second embodiment. Figure 2 The system in the embodiment is compared with Figure 1 The difference in the embodiment is that the risk profile evaluation and risk decision-making system further includes a question-and-answer module. The question-and-answer module can be connected to the data acquisition module, and the question-and-answer module is also connected to the risk decision-making module.

[0073] Among them, the question-and-answer module is used to obtain a user query request. The user query request carries query location information and a query event. When the query event is included in the target risk profile, the response strategy for the query event in the target risk profile is output. When the query event is not included in the target risk profile, an indication information that the query event does not exist at the query location is output.

[0074] In some embodiments, the risk profile evaluation and risk decision-making system may include a display module. The display module can display a question-and-answer page. The user inputs query content through the question-and-answer page, and the display module generates a user query request in response to the query content input by the user. Among them, the user query request may include the query content input by the user.

[0075] In some embodiments, the data acquisition module may, in response to a user query request, obtain the multi-source data that matches the location information carried in the user query request from the database, and the Q&A module may determine the query type of the query content and output a corresponding answer on the Q&A page according to the query type of the query content.

[0076] For example, in the case where the query type is a risk decision type, the coping strategies of the target risk profile and potential future risks are output on the Q&A page. For example, in the case where the query type is a risk assessment type, the target risk profile is output on the Q&A page. For another example, in the case where the query type is an information query type, keywords are obtained from the query content input by the user, and solution information that matches the keywords is obtained from a pre-constructed knowledge graph, and the solution information is output on the Q&A page.

[0077] Figure 3 It is a schematic structural diagram of a risk profile evaluation and risk decision-making system provided for the third embodiment. Figure 3 The system in the embodiment is compared with Figure 1 The difference in the embodiment is that the risk profile evaluation and risk decision-making system further includes a voice broadcast module. The voice broadcast module can be connected to the risk profile evaluation module.

[0078] The voice broadcast module is used to determine the risk broadcast information corresponding to the target risk profile, obtain the type of the computer device with the system, determine at least one language type for broadcasting according to the type of the computer device, convert the risk broadcast information into the languages of the respective language types to obtain respective target play information, and perform voice broadcast on the respective target play information.

[0079] Among them, the type of the computer device can be a type for personal use, a type for public place warning, a type for enterprise use, etc. Exemplarily, in the case where the type of the computer device is a type for personal use, the language type for broadcasting can be determined according to the language habit of the person. For another example, in the case where the type of the computer device is a type for public place warning, the language types for broadcasting are determined to be Chinese and English, etc. For another example, in the case where the type of the computer device is a type for enterprise use, the language type for broadcasting is the language matched by the enterprise.

[0080] In some embodiments, determining the risk broadcast information corresponding to the target risk profile can be achieved through the following method: obtaining the location information and / or participant information of the risk event in the target risk profile, generating risk description information according to the coping strategy of the target risk profile and according to the location information and / or participant information of the risk event, and converting the risk description information into voice-type information to obtain the risk broadcast information.

[0081] In some embodiments, when the type of the computer device is a public place warning type, the target risk level can also be determined according to the target risk profile, and the corresponding loudspeaker can be determined according to the target risk level, and each target playback data can be sent to the corresponding loudspeaker so that the corresponding loudspeaker can perform voice broadcast.

[0082] Figure 4 FIG. 4 is a schematic structural diagram of a risk profile evaluation and risk decision-making system provided for the fourth embodiment. Figure 4 The system in the embodiment is compared with Figure 1 The embodiment is different in that the risk profile evaluation and risk decision-making system further includes a question-and-answer module and a voice broadcast module.

[0083] Figure 5 FIG. 5 is a schematic structural diagram of a risk profile evaluation and risk decision-making system provided for the fifth embodiment. Figure 5 The system in the embodiment is compared with Figure 1 The embodiment is different in that the data preprocessing module includes: a data cleaning sub-module, a data format conversion sub-module, and a data standardization sub-module.

[0084] The data cleaning sub-module is used to perform image duplicate removal processing and image denoising processing on the initial image data to obtain preset image data, perform text cleaning processing on the initial text data to obtain preset text data, and perform data cleaning processing on the initial sensing data to obtain preset sensing data;

[0085] The data format conversion sub-module is used to perform image format conversion on the preset image data to obtain set image data in a set image format, perform text format conversion on the preset text data to obtain set text data in a set text format, and perform data format conversion on the preset sensing data to obtain set sensing data in a set data format;

[0086] The data standardization sub-module is used to perform image size adjustment processing on the set image data to obtain target image data with a preset image size, perform standardization processing on the set text data to obtain target text data, and perform standardization processing on the set sensing data to obtain target sensing data with a preset dimension and a preset range.

[0087] Exemplarily, performing image duplicate removal processing and image denoising processing on the initial image data to obtain preset image data can be achieved by the following method: performing image duplicate removal processing on the initial image data to obtain the de-duplicated image, and performing image denoising processing on the de-duplicated image to obtain preset image data.

[0088] Exemplarily, the initial text data is subjected to text cleaning processing to obtain preset text data, which can be achieved by the following method: removing special characters, Hyper Text Markup Language (HTML) tags, stop words, etc. from the initial text data to obtain preset text data.

[0089] Exemplarily, the initial sensing data is subjected to data cleaning processing to obtain preset sensing data, which can be achieved by the following method: removing noise and outliers from the initial sensing data to obtain preset sensing data.

[0090] Exemplarily, the preset image size can be the preset image pixel size.

[0091] Exemplarily, the set text data is subjected to standardization processing to obtain target text data, which can be achieved by the following method: converting the set text data into unified terms and formats and matching different expressions of the same entity to obtain target text data.

[0092] Exemplarily, the standardization processing of the set sensing data can be achieved by the following method: performing dimension conversion processing and normalization processing on the set sensing data.

[0093] Figure 6 It is a schematic structural diagram of a risk portrait evaluation and risk decision-making system provided for the sixth embodiment. Figure 6 The system in the embodiment is compared with Figure 1 The difference from the embodiment is that the risk portrait evaluation module includes: a risk portrait evaluation sub-module and a risk portrait fusion sub-module.

[0094] The risk portrait evaluation sub-module is used to process the target image data by using an image risk portrait evaluation model to obtain an image risk portrait, process the target text data by using a text risk portrait evaluation model to obtain a text risk portrait, and preprocess the target sensing data by using a sensing risk portrait evaluation model to obtain a sensing risk portrait.

[0095] The risk portrait fusion sub-module is used to fuse the image risk portrait, the text risk portrait, and the sensing risk portrait to obtain a target risk portrait.

[0096] In some embodiments, the risk portrait evaluation sub-module includes: an image risk portrait evaluation unit, a text risk portrait evaluation unit, and a sensing risk portrait evaluation unit.

[0097] An image risk profile assessment unit, configured to determine at least two image risk assessment dimensions, and adopt respective image feature extraction modules corresponding to the image risk assessment dimensions to respectively extract features from the target image data to obtain respective image feature information, use an image feature assessment model to process the respective image feature information to obtain assessment results under the respective image risk assessment dimensions, and determine the image risk profile according to the assessment results under the respective image risk assessment dimensions;

[0098] A text risk profile assessment unit, configured to determine at least two text risk assessment dimensions, and adopt respective text feature extraction modules corresponding to the text risk assessment dimensions to respectively extract features from the target text data to obtain respective text feature information, use a text feature assessment model to process the respective text feature information to obtain assessment results under the respective text risk assessment dimensions, and determine the text risk profile according to the assessment results under the respective text risk assessment dimensions;

[0099] A sensing risk profile assessment unit, configured to determine at least two sensing risk assessment dimensions, and adopt respective sensing feature extraction modules corresponding to the sensing risk assessment dimensions to respectively extract features from the target sensing data to obtain respective sensing feature information, use a sensing feature assessment model to process the respective sensing feature information to obtain assessment results under the respective sensing risk assessment dimensions, and determine the sensing risk profile according to the assessment results under the respective sensing risk assessment dimensions.

[0100] Exemplarily, at least two image risk assessment dimensions, at least two text risk assessment dimensions, and at least two sensing risk assessment dimensions may be the same or may be partially different.

[0101] Exemplarily, use an image feature assessment model to process the respective image feature information to obtain probability values under the respective image risk assessment dimensions, and determine the assessment results under the respective image risk assessment dimensions according to the probability values under the respective image risk assessment dimensions.

[0102] Exemplarily, use a text feature assessment model to process the respective text feature information to obtain probability values under the respective text risk assessment dimensions, and determine the assessment results under the respective text risk assessment dimensions according to the probability values under the respective text risk assessment dimensions.

[0103] Exemplarily, use a sensing feature assessment model to process the respective sensing feature information to obtain probability values under the respective sensing risk assessment dimensions, and determine the assessment results under the respective sensing risk assessment dimensions according to the probability values under the respective sensing risk assessment dimensions.

[0104] Exemplarily, the evaluation result can be a score, with a value in the range of 0 - 100. In this way, the probability value under each of the sensing risk assessment dimensions can be multiplied by 100 to obtain the evaluation result under each of the sensing risk assessment dimensions. Also exemplarily, the evaluation result can be a probability value, and the probability value under each of the sensing risk assessment dimensions is determined as the evaluation result under each of the sensing risk assessment dimensions.

[0105] In some embodiments, the risk profile fusion sub-module includes: an evaluation result fusion unit and a risk profile fusion unit.

[0106] The evaluation result fusion unit is configured to fuse the evaluation results under each of the image risk assessment dimensions, the evaluation results under each of the text risk assessment dimensions, and the evaluation results under each of the sensing risk assessment dimensions to obtain the evaluation results under multiple comprehensive risk assessment dimensions;

[0107] The risk profile fusion unit is configured to determine the weights under each of the comprehensive risk assessment dimensions according to the weights under each of the image risk assessment dimensions, the weights under each of the text risk assessment dimensions, and the weights under each of the sensing risk assessment dimensions obtained, and perform weighted fusion on the evaluation results under each of the comprehensive risk assessment dimensions and the weights under each of the comprehensive risk assessment dimensions to obtain the target risk profile.

[0108] Exemplarily, the evaluation results under each of the image risk assessment dimensions include that the evaluation result of dimension A1 is X1, and the evaluation result of dimension A2 is X2 (X1 + X2 equals 1, or X1 + X2 equals the total score value). The evaluation results under each of the text risk assessment dimensions include that the evaluation result of dimension A3 is X3, and the evaluation result of dimension A4 is X4 (X3 + X4 equals 1, or X3 + X4 equals the total score value). The evaluation results under each of the sensing risk assessment dimensions include that the evaluation result of dimension A2 is X5, and the evaluation result of dimension A3 is X6 (X5 + X6 equals 1, or X5 + X6 equals the total score value). Then, the evaluation results under multiple comprehensive risk assessment dimensions include that the evaluation result of dimension A1 is X1 / (X1 + X2 + X3 + X4 + X5 + X6), the evaluation result of dimension A2 is (X2 + X5) / (X1 + X2 + X3 + X4 + X5 + X6), the evaluation result of dimension A3 is (X3 + X6) / (X1 + X2 + X3 + X4 + X5 + X6), and the evaluation result of dimension A4 is X4 / (X1 + X2 + X3 + X4 + X5 + X6).

[0109] Exemplarily, the weights under each of the image risk assessment dimensions include the weight of dimension B1 being Y1, and the weight of dimension B2 being Y2 (where Y1 + Y2 equals 1). The weights under each of the text risk assessment dimensions include the weight of dimension B3 being Y3, and the weight of dimension B4 being Y4 (where Y3 + Y4 equals 1). The weights under each of the sensing risk assessment dimensions include the weight of dimension B2 being Y5, and the weight of dimension B3 being Y6 (where Y5 + Y6 equals 1). Then, the weights under each of the comprehensive risk assessment dimensions include the weight of dimension B1 being Y1 / (Y1 + Y2 + Y3 + Y4 + Y5 + Y6), the weight of dimension B2 being (Y2 + Y5) / (Y1 + Y2 + Y3 + Y4 + Y5 + Y6), the weight of dimension B3 being (Y3 + Y6) / (Y1 + Y2 + Y3 + Y4 + Y5 + Y6), and the weight of dimension B4 being Y4 / (Y1 + Y2 + Y3 + Y4 + Y5 + Y6).

[0110] Exemplarily, the weighted fusion of the evaluation results under each of the comprehensive risk assessment dimensions and the weights under each of the comprehensive risk assessment dimensions can be weighted summation.

[0111] Figure 7 It is a schematic structural diagram of a risk portrait evaluation and risk decision-making system provided for the seventh embodiment. Figure 7 The system in the embodiment is compared with Figure 1 The difference in the embodiment is that the risk decision-making module includes: a rule inference sub-module and a probability inference sub-module.

[0112] The rule inference sub-module is used to input the target risk portrait and the multi-source data into the rule inference model, and output the coping strategy for the target risk portrait through the rule inference model.

[0113] The probability inference sub-module is used to input the multi-source data, historical multi-source data, and the target risk portrait into the probability inference model, and output the future potential risks through the probability inference model.

[0114] Exemplarily, the rule inference sub-module can be a module based on an expert system. Exemplarily, the probability inference sub-module can be a module based on a decision tree.

[0115] An expert system is a type of computer intelligent information system with a large amount of specialized knowledge. It uses the specialized knowledge in a specific field and reasoning techniques in artificial intelligence to solve and simulate various complex and specific problems that usually require human experts to solve. An expert system generally consists of four main parts: a knowledge base, an inference engine, a user interface, and an explanation system. Among them, the knowledge base stores the knowledge and experience of domain experts, and the inference engine is responsible for reasoning based on the input problems using the knowledge in the knowledge base to obtain solutions. The user interface allows users to interact with the expert system, while the explanation system is used to explain the behavior and decision-making basis of the expert system. Rule-based reasoning is one of the most commonly used reasoning methods in expert systems. It formulates rules based on the experience and knowledge accumulated by experts in the field, and the inference engine traverses all the rules in the knowledge base to determine whether the conditions of the rules are met, so as to execute corresponding actions.

[0116] A decision tree is a method for approximating discrete function values based on the known probabilities of various situations occurring; it is a graphical method for evaluating project risks and judging the feasibility of decisions, and is an intuitive application of probability analysis.

[0117] Figure 8 A schematic structural diagram of a risk portrait evaluation and risk decision-making system provided for the eighth embodiment Figure 8 The system in the embodiment is compared with Figure 1 The difference between the embodiments is that the risk portrait evaluation and risk decision-making system further includes a data prediction module; the data prediction module is connected to the data preprocessing module, and the data prediction module is also connected to the data acquisition module.

[0118] The data prediction module is used to predict the initial image data using an image prediction model to obtain predicted image data, predict the initial text data using a text prediction model to obtain predicted text data, and predict the initial sensing data using a sensing prediction model to obtain predicted sensing data;

[0119] The data preprocessing module is used to preprocess the predicted image data and the initial image data to obtain the target image data, preprocess the predicted text data and the initial text data to obtain the target text data, and preprocess the predicted sensing data and the initial sensing data to obtain the target sensing data.

[0120] The embodiment of the present application proposes a multi-dimensional information fusion personal safety portrait and intelligent decision-making system (i.e., the above-mentioned risk portrait evaluation and risk decision-making system). By integrating multiple data sources and applying advanced data processing and analysis technologies, it can identify and evaluate potential personal safety risks in real time and provide scientific and reasonable decision-making support for managers.

[0121] Figure 9 Schematic diagram of the structure of a multi-dimensional information fusion personal safety portrait and intelligent decision-making system provided for some embodiments, as follows Figure 9 shown, the system includes a data acquisition module, a data preprocessing module, a personal risk intelligent identification module (corresponding to the above-mentioned risk portrait evaluation module), an intelligent question and answer module (corresponding to the above-mentioned question and answer module), a decision support module (corresponding to the above-mentioned risk decision module), and a language intelligent broadcast module (corresponding to the above-mentioned voice broadcast module).

[0122] Among them, the data acquisition module is used to collect personal safety-related data from multiple data sources; the data preprocessing module is used to clean, format, and standardize the collected data to ensure that the data quality meets the requirements of subsequent processing; the personal risk intelligent identification module is used to analyze the preprocessed data through machine learning algorithms, identify potential personal safety risks, and generate risk portraits; the intelligent question and answer module is used to provide real-time question and answer services according to user query requests through natural language processing technology to help users quickly obtain safety information and risk response measures; the decision support module is used to provide decision-making suggestions for managers according to risk portraits and real-time data by using expert systems and decision tree intelligent decision-making technologies to prevent and respond to potential risks; the voice intelligent broadcast module is used to notify relevant personnel in real time of the identified key risk information through voice broadcast to ensure the timely transmission of information.

[0123] As a further improvement of the embodiments of the present application, the data acquisition module includes, but is not limited to, video surveillance data, sensor data, social media data, public safety databases, and data uploaded by user personal devices.

[0124] As a further improvement of the embodiments of the present application, the data preprocessing module includes a data cleaning sub-module, a data formatting sub-module, and a data standardization sub-module.

[0125] As a further improvement of the embodiments of the present application, the data cleaning sub-module is responsible for removing noise and outliers in the data to ensure data accuracy; the data formatting sub-module converts data from different sources into a unified format for subsequent processing; the data standardization sub-module converts the data into standard dimensions and ranges to eliminate the dimensional differences between different data sources and facilitate subsequent analysis and processing.

[0126] As a further improvement of the embodiments of the present application, the personal risk intelligent identification module uses deep learning technology. By constructing a neural network model, it can efficiently process and analyze large-scale data; and can identify security risks at the individual level, and can also analyze group behavior patterns, so as to predict and discover potential public security problems and generate risk portraits; at the same time, through continuous learning and optimization, it can improve the accuracy and timeliness of risk identification, providing strong technical support for security protection.

[0127] As a further improvement of the embodiments of the present application, the intelligent question-answering module includes a natural language understanding sub-module and a natural language generation sub-module.

[0128] As a further improvement of the embodiments of the present application, the natural language understanding sub-module is responsible for parsing the user's query request, extracting key information, and understanding the user's intention; the natural language generation sub-module then generates accurate and fluent answer content according to the analysis result, ensuring that users can obtain clear and easy-to-understand security information and risk response measures.

[0129] As a further improvement of the embodiments of the present application, the decision support module uses a rule-based inference engine and probabilistic inference technology, which can combine risk portraits and real-time data for multi-dimensional analysis and evaluation; and can provide response strategies for specific risks, and can also predict potential risks that may occur in the future based on historical data and current situations, thus providing comprehensive decision support for management personnel.

[0130] The beneficial effects of the embodiments of the present application include:

[0131] Improve the intelligence and comprehensiveness of security monitoring: By integrating multi-dimensional data sources such as video monitoring, sensors, and social media, and combining advanced data processing and analysis technologies, the embodiments of the present application can achieve comprehensive monitoring in complex scenarios and improve the intelligence level of security monitoring. This not only overcomes the problems of missed detection and misjudgment caused by human fatigue and distraction in traditional monitoring methods, but also enables the system to process and analyze large-scale data and more accurately identify potential personal security risks.

[0132] Provide scientific and timely decision support: The built-in decision support module of the embodiments of the present application uses expert systems and decision tree intelligent decision-making technologies, which can provide scientific and reasonable decision-making suggestions for management personnel based on risk portraits and real-time data. This helps management personnel to quickly make correct judgments and response measures in the face of complex and changeable security situations, thus effectively preventing and coping with potential risks.

[0133] Enhancing User Experience and Information Transmission Efficiency: Through the intelligent Q&A module and the voice intelligent broadcast module, the embodiments of this application not only provide users with a convenient way to query security information, but also, through voice broadcast, notify relevant personnel of key risk information in real time. This multi-channel information transmission method not only improves the user experience, but also ensures the timeliness and effectiveness of information, enabling users with different language backgrounds to receive key security information in a timely manner, thereby enhancing the overall security protection ability.

[0134] The embodiments of this application provide a multi-dimensional information fusion personal safety portrait and intelligent decision-making system, which consists of a data collection module, a data preprocessing module, a personal risk intelligent identification module, an intelligent Q&A module, a decision support module, and a language intelligent broadcast module. The data collection module is used to collect personal safety-related data from multiple data sources; the data preprocessing module is used to clean, format, and standardize the collected data to ensure that the data quality meets the requirements of subsequent processing; the personal risk intelligent identification module is used to analyze the preprocessed data through machine learning algorithms, identify potential personal safety risks, and generate risk portraits; the intelligent Q&A module is used to provide real-time Q&A services according to user query requests through natural language processing technology to help users quickly obtain security information and risk response measures; the decision support module is used to provide decision-making suggestions for managers according to risk portraits and real-time data by using expert systems and decision tree intelligent decision-making technologies to prevent and respond to potential risks; the voice intelligent broadcast module is used to notify relevant personnel of the identified key risk information in real time through voice broadcast to ensure the timely transmission of information.

[0135] In the embodiments of this application, the data collection module includes but is not limited to video surveillance data, sensor data, social media data, public security databases, and data uploaded by user personal devices.

[0136] In the embodiments of this application, the data preprocessing module includes a data cleaning sub-module, a data formatting sub-module, and a data standardization sub-module. The data cleaning sub-module is responsible for removing noise and outliers in the data to ensure data accuracy; the data formatting sub-module converts data from different sources into a unified format for subsequent processing; the data standardization sub-module converts the data into standard dimensions and ranges to eliminate the dimensional differences between different data sources and facilitate subsequent analysis and processing.

[0137] The personal risk intelligent identification module adopts deep learning technology. By constructing a neural network model, it can efficiently process and analyze large-scale data. It can identify security risks at the individual level and analyze group behavior patterns to predict and discover potential public safety issues. At the same time, through continuous learning and optimization, it can improve the accuracy and timeliness of risk identification, providing strong technical support for security protection.

[0138] The intelligent Q&A module includes a natural language understanding sub-module and a natural language generation sub-module. The natural language understanding sub-module is responsible for parsing the user's query request, extracting key information, and understanding the user's intention. The natural language generation sub-module then generates accurate and fluent answer content based on the analysis results to ensure that users can obtain clear and easy-to-understand security information and risk response measures.

[0139] The decision support module adopts a rule-based inference engine and probabilistic inference technology. It can combine the risk profile and real-time data for multi-dimensional analysis and evaluation. It can also provide response strategies for specific risks and predict potential risks that may occur in the future based on historical data and current situations, thus providing comprehensive decision support for managers.

[0140] The voice intelligent broadcast module has a multi-language support function. It can translate and broadcast risk information into the corresponding language according to the user-set or automatically recognized language preference, ensuring that users with different language backgrounds can receive key security information in a timely manner.

[0141] Taking the power system as an example to illustrate the implementation process of the embodiments of the present application:

[0142] In the power system, the multi-dimensional information fusion personal safety profile and intelligent decision-making system of the embodiments of the present application can play an important role. By integrating multi-dimensional data sources such as video monitoring, sensor data, operation records, and meteorological information of the power system, the system can monitor the safety status of power facilities in real time and conduct intelligent identification and early warning of potential risks.

[0143] The data acquisition module collects real-time image data from the video monitoring systems at key locations such as substations, transmission lines, and distribution rooms, and conducts comprehensive analysis in combination with sensor data (such as temperature, humidity, current, voltage, etc.). In addition, the system can also access the personal device data of operators, such as the sensors on the smart watches or safety helmets they wear, to monitor the physiological status and location information of operators in real time.

[0144] The data preprocessing module cleans, formats and standardizes the collected data to ensure that the data quality meets the requirements of subsequent processing. For example, the data cleaning submodule removes noise and outliers, the data formatting submodule converts data from different sources into a unified format, and the data standardization submodule converts data into standard dimensions and ranges.

[0145] The personal risk intelligent identification module builds a deep learning model to efficiently process and analyze large-scale data, identifying individual safety risks and group behavior patterns. For example, the system can identify potential risks such as operators not wearing safety helmets or not operating according to regulations near high-voltage equipment, and generate corresponding risk profiles.

[0146] The intelligent question-and-answer module provides real-time question-and-answer services for power system managers and operators. For example, when operators encounter safety issues during operations, they can query relevant safety operating procedures and response measures through the intelligent question-and-answer module to ensure operational safety.

[0147] The decision support module combines risk profiling and real-time data, and uses expert systems and decision tree intelligent decision-making technology to provide managers with scientific and reasonable decision-making suggestions. For example, under severe weather conditions, the system can predict potential power facility failure risks based on meteorological information and real-time data of power facilities, and provide managers with suggestions for countermeasures.

[0148] The voice intelligent broadcast module notifies relevant personnel of the identified key risk information in real time through voice broadcast to ensure timely communication of information. For example, when an abnormal situation is found near a power transmission line, the system can automatically translate the warning information into the language set by the operator and broadcast it through the on-site loudspeaker to remind relevant personnel to take countermeasures.

[0149] Through the implementation of the embodiments of the present application, the power system can achieve comprehensive monitoring in complex scenarios, improve the intelligence level of security monitoring, provide scientific and timely decision-making support for managers, enhance user experience and information communication efficiency, thereby effectively preventing and responding to potential risks and ensuring the safe and stable operation of the power system.

[0150] The implementation process of the embodiment of this application is described by taking the security monitoring of a large shopping mall as an application:

[0151] The multi-dimensional information fusion personal safety portrait and intelligent decision-making system was applied to the security monitoring of a large shopping mall. Multiple high-definition video surveillance cameras were installed in the shopping mall to monitor the flow and activities of people in the mall in real time. In addition, various sensors such as smoke detectors, temperature detectors and emergency buttons were deployed to collect data related to personal safety.

[0152] The data acquisition module integrates multi-source data such as video surveillance data, sensor data, and social media data. For example, by analyzing social media data, the system can identify external events that may affect the safety of the shopping mall, such as emergencies in the nearby area or public comments on the safety status of the shopping mall.

[0153] The data preprocessing module cleans, formats, and standardizes the collected data. For example, the data cleaning sub-module automatically removes irrelevant background information in the video surveillance data, such as billboards and decorations, to improve the accuracy of subsequent analysis. The data formatting sub-module converts data from different sources into a unified format for subsequent processing. The data standardization sub-module converts the data into standard dimensions and ranges to ensure compatibility between different data sources.

[0154] The personal risk intelligent identification module analyzes the preprocessed data through the constructed neural network model to identify potential personal safety risks. For example, by analyzing video surveillance data, the system can identify abnormal behaviors in the mall, such as fighting and stealing, and generate corresponding risk portraits.

[0155] The intelligent question-answering module provides real-time question-answering services for users. For example, when a user queries "Is there a fire alarm in the mall?" through a mobile device, the natural language understanding sub-module parses the user's query request and extracts key information. The natural language generation sub-module generates the answer content based on the analysis result: There is currently no fire alarm in the mall, but please pay attention to the emergency exit signs for a quick evacuation in case of an emergency.

[0156] The decision support module provides decision-making suggestions for management personnel based on the risk portraits and real-time data, using expert systems and decision tree intelligent decision-making technologies. For example, when the system detects that the personnel density in a certain area of the mall exceeds the safety threshold, the decision support module will recommend that management personnel take evacuation measures to prevent crowded stampedes.

[0157] The voice intelligent broadcast module notifies relevant personnel of the identified key risk information in real time through voice broadcasts. For example, when the system detects a fire in the mall, the voice intelligent broadcast module will automatically switch to the multi-language broadcast mode and translate and broadcast the fire alarm information into the corresponding language according to the user's set language preference to ensure that all personnel can receive key safety information in a timely manner.

[0158] Through the implementation of the embodiments of this application, the safety monitoring level of this large shopping mall has been significantly improved. Management personnel can make more scientific and timely decisions, effectively prevent and respond to potential risks, thereby providing a safer shopping and working environment for customers and employees.

[0159] In summary, a multi-dimensional information fusion personal safety portrait and intelligent decision-making system according to an embodiment of the present application can comprehensively monitor complex scenarios through multi-dimensional information fusion technology, improve the intelligent level of security monitoring, provide scientific and timely decision-making support for management personnel, enhance the user experience and information transmission efficiency, thereby effectively preventing and responding to potential risks and ensuring the safe and stable operation of the system.

[0160] Each module in the above risk portrait assessment and risk decision-making system can be implemented in whole or in part by software, hardware, and their combinations. The above-mentioned modules can be embedded in the processor of the computer device in hardware form or be independent of it, or be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each of the above modules.

[0161] In an exemplary embodiment, Figure 10 FIG. is a schematic structural diagram of a computer device provided for some embodiments. The computer device includes the risk portrait assessment and risk decision-making system according to any one of the above.

[0162] In some embodiments, the computer device may further include a processor, a memory, an input / output interface, a communication interface, and an input device. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface and the input device are connected to the system bus through the input / output interface. Among them, 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 input / output interface of the computer device is used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communicating with external terminals in a wired or wireless manner. The wireless manner can be achieved through Wireless Fidelity (WIFI), a mobile cellular network, Near Field Communication (NFC), or other technologies. The input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the computer device housing, or an external keyboard, touchpad, or mouse, etc.

[0163] Those skilled in the art can understand that Figure 10 the structure shown in FIG. is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0164] The processor, each functional module, or each functional unit in any embodiment of the present application may include any one or more of the following integrations: general-purpose processor, application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), central processing unit (CPU), graphics processing unit (GPU), embedded neural-network processing units (NPU), controller, microcontroller, microprocessor, programmable logic device, discrete gate or transistor logic device, discrete hardware component, quantum computing-based data processing logic unit, artificial intelligence (AI) processor, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0165] The memory or computer-readable storage medium in any embodiment of the present application may include at least one of non-volatile memory and volatile memory. The non-volatile memory includes the integration of one or more of the following: Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Ferromagnetic Random Access Memory (FRAM), Flash Memory, magnetic surface memory, optical disc, Compact Disc Read-Only Memory (CD-ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, Resistive Random Access Memory (ReRAM), Magnetoresistive Random Access Memory (MRAM), Ferroelectric Random Access Memory (FRAM), Phase Change Memory (PCM), graphene memory, volatile memory, etc. The volatile memory includes the integration of one or more of the following: Random Access Memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), etc.

[0166] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in the present application.

[0167] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation to the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A risk profiling assessment and risk decision-making system, characterized in that, The system includes: A data acquisition module for obtaining multi-source data from a database; A data preprocessing module for obtaining initial image data, initial text data, and initial sensing data from the multi-source data, preprocessing the initial image data to obtain target image data, preprocessing the initial text data to obtain target text data, and preprocessing the initial sensing data to obtain target sensing data; A risk profile assessment module for processing the target image data using an image risk profile assessment model to obtain an image risk profile, processing the target text data using a text risk profile assessment model to obtain a text risk profile, preprocessing the target sensing data using a sensing risk profile assessment model to obtain a sensing risk profile, and fusing the image risk profile, the text risk profile, and the sensing risk profile to obtain a target risk profile; A risk decision-making module for determining a coping strategy for the target risk profile and potential future risks based on the target risk profile, the multi-source data, and historical multi-source data.

2. The system according to claim 1, wherein The system further includes: A question-and-answer module for obtaining a user query request, where the user query request carries query location information and a query event, and outputting a coping strategy for the query event in the target risk profile if the query event is included in the target risk profile, or outputting an indication information that the query event does not exist at the query location if the query event is not included in the target risk profile.

3. The system according to claim 1, characterized in that, The system further includes: A voice broadcast module for determining risk broadcast information corresponding to the target risk profile, obtaining the type of a computer device having the system, determining at least one language type for broadcast according to the type of the computer device, converting the risk broadcast information using the languages of the respective language types to obtain respective target play information, and performing voice broadcast on the respective target play information.

4. The system according to claim 1, wherein The data preprocessing module includes: A data cleaning sub-module for performing image duplicate removal processing and image denoising processing on the initial image data to obtain preset image data, performing text cleaning processing on the initial text data to obtain preset text data, and performing data cleaning processing on the initial sensing data to obtain preset sensing data; A data format conversion sub-module for performing image format conversion on the preset image data to obtain set image data in a set image format, performing text format conversion on the preset text data to obtain set text data in a set text format, and performing data format conversion on the preset sensing data to obtain set sensing data in a set data format; A data standardization sub-module for performing image size adjustment processing on the set image data to obtain target image data with a preset image size, performing standardization processing on the set text data to obtain target text data, and performing standardization processing on the set sensing data to obtain target sensing data with a preset dimension and a preset range.

5. The system according to any one of claims 1 to 4, characterized in that, The risk profile assessment module includes: The risk profile assessment sub-module is used to process the target image data using an image risk profile assessment model to obtain an image risk profile, process the target text data using a text risk profile assessment model to obtain a text risk profile, and preprocess the target sensing data using a sensing risk profile assessment model to obtain a sensing risk profile; The risk profile fusion sub-module is used to fuse the image risk profile, the text risk profile, and the sensing risk profile to obtain a target risk profile.

6. The system according to claim 5, wherein The risk profile assessment sub-module includes: The image risk profile assessment unit is used to determine at least two image risk assessment dimensions, use each image feature extraction module corresponding to each image risk assessment dimension to extract features from the target image data respectively to obtain each image feature information, use an image feature assessment model to process each image feature information to obtain the evaluation results under each image risk assessment dimension, and determine the image risk profile according to the evaluation results under each image risk assessment dimension; The text risk profile assessment unit is used to determine at least two text risk assessment dimensions, use each text feature extraction module corresponding to each text risk assessment dimension to extract features from the target text data respectively to obtain each text feature information, use a text feature assessment model to process each text feature information to obtain the evaluation results under each text risk assessment dimension, and determine the text risk profile according to the evaluation results under each text risk assessment dimension; The sensing risk profile assessment unit is used to determine at least two sensing risk assessment dimensions, use each sensing feature extraction module corresponding to each sensing risk assessment dimension to extract features from the target sensing data respectively to obtain each sensing feature information, use a sensing feature assessment model to process each sensing feature information to obtain the evaluation results under each sensing risk assessment dimension, and determine the sensing risk profile according to the evaluation results under each sensing risk assessment dimension.

7. The system according to claim 6, characterized in that, The risk profile fusion sub-module includes: The evaluation result fusion unit is used to fuse the evaluation results under each image risk assessment dimension, the evaluation results under each text risk assessment dimension, and the evaluation results under each sensing risk assessment dimension to obtain the evaluation results under multiple comprehensive risk assessment dimensions; The risk profile fusion unit is used to determine the weights under each comprehensive risk assessment dimension according to the weights under each image risk assessment dimension, the weights under each text risk assessment dimension, and the weights under each sensing risk assessment dimension obtained, and perform weighted fusion on the evaluation results under each comprehensive risk assessment dimension and the weights under each comprehensive risk assessment dimension to obtain the target risk profile.

8. The system according to any one of claims 1 to 4, characterized in that The risk decision module includes: The rule inference sub-module is used to input the target risk profile and the multi-source data into the rule inference model, and output the coping strategy of the target risk profile through the rule inference model; A probability reasoning sub-module, configured to input the multi-source data, historical multi-source data, and the target risk profile into a probability reasoning model, and output the future potential risks through the probability reasoning model.

9. The system according to any one of claims 1 to 4, characterized in that, The system further includes a data prediction module: The data prediction module is configured to use an image prediction model to predict the initial image data to obtain predicted image data, use a text prediction model to predict the initial text data to obtain predicted text data, and use a sensing prediction model to predict the initial sensing data to obtain predicted sensing data; A data preprocessing module, configured to preprocess the predicted image data and the initial image data to obtain the target image data, preprocess the predicted text data and the initial text data to obtain the target text data, and preprocess the predicted sensing data and the initial sensing data to obtain the target sensing data.

10. A computer device, characterized in that, The computer device includes the risk profile evaluation and risk decision-making system according to any one of claims 1 to 9.

Images