An intelligent risk prediction and identification system, equipment and device based on big data graph computing

Through an intelligent risk prediction and identification system based on big data map calculation, the problem that existing risk control systems cannot identify and respond to new risks is solved, and more accurate and flexible risk assessment and control are achieved, reducing corporate interests losses.

CN113344460BActive Publication Date: 2024-12-06SHANGHAI SOFTLINE INFORMATION TECH
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202110769450.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-07-07
Publication Date
2024-12-06
Estimated Expiration
2041-07-07

AI Technical Summary

Technical Problem

The existing risk control system cannot effectively identify and respond to new risks, especially when facing multiple influencing factors, which leads to the inability to update the risk database in a timely manner and the inability to flexibly respond to risk assessments of different customers and products, resulting in loss of corporate interests.

Method used

Using an intelligent risk prediction and identification system based on big data map calculation, by obtaining multiple case instances, data screening and risk feature extraction, a risk feature library model is constructed, and the target case is analyzed, and the risk characteristics and their associations are displayed. Relationships to achieve continuous optimization and update of the model.

Benefits of technology

It improves the accuracy and efficiency of risk identification, can provide flexible processing methods based on different influencing factors, reduce corporate profit losses, adapt to industry development, and provide more accurate risk assessment and control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113344460B_ABST
    Figure CN113344460B_ABST
Patent Text Reader

Abstract

The present invention discloses an intelligent risk prediction and identification system, equipment and device based on big data graph calculation, which belongs to the field of big data analysis technology and knowledge graph technology. It includes analyzing and filtering multiple historical case instance data information, extracting their risk characteristics, calculating feature weights and similarities between cases, establishing a risk feature library model, using the risk feature library model to analyze and calculate the target case, displaying the obtained data information and risk characteristics, and displaying the correlation relationship and path between them. The present invention can analyze and find out the hidden risks in enterprise cases, and has the advantages of simple structure, reasonable design and easy manufacturing.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of big data analysis technology and knowledge graph technology, and specifically, relates to an intelligent risk prediction and identification system, equipment and device based on big data graph calculation. Background Art

[0002] The current risk control system within financial enterprises relies too much on the professional capabilities of business personnel. The internal risk control system of the enterprise can only detect single or old past risks. When facing new risks, it cannot update the risk database in time. Moreover, the same risk may cause different results after being mixed with different influencing factors, which may ultimately cause serious loss of profits to the enterprise.

[0003] Moreover, most of the existing risk control systems can only monitor and provide feedback on dangerous data information entered into the database. The intelligent AI learning level of the risk control system is not enough. When sensitive information outside the database appears, it cannot be judged well and effectively, and manual database upgrades are required on a regular basis. Therefore, if the risk problem encountered by the enterprise does not exist in the database, the risk control system cannot detect it and warn the enterprise in the first time, and the interests of the enterprise will be greatly affected. In addition, the risk control system cannot provide the best solution to solve the troubles encountered, which may eventually cause unpredictable losses.

[0004] At the same time, when facing different customers, product channels, and handling employees, these factors will have an impact on risk assessment, thereby affecting the assessment results of the risk control system. When facing the same risk problem, if the above factors are different, the risk control system may not be able to make a judgment and still use fixed assessment methods and solutions. It cannot flexibly provide different treatment methods according to different influencing factors. The resulting results are likely to be very bad, and ultimately cause significant losses to the company's interests. Summary of the invention

[0005] 1. Problems to be solved

[0006] In view of the problem that the existing risk control system can only identify the impact and risk of a single event on business operations, but cannot timely update the scope and degree of impact of different risks on business operations according to the development of the industry, the present invention provides an intelligent risk prediction and identification system, equipment and device based on big data graph calculation.

[0007] 2. Technical solution

[0008] To solve the above problems, the present invention adopts the following technical solutions.

[0009] An intelligent risk prediction and identification system based on big data graph computing adopts the following steps:

[0010] Step 1: Obtain multiple case instances, perform rough classification, and filter the data information in the case instances;

[0011] Step 2: Extract risk features for each case instance based on the processing results of step 1, and calculate the risk feature weight and the similarity of each case instance;

[0012] Step 3: Construct a risk feature library model based on the risk feature weights and case instance similarities calculated in step 2;

[0013] Step 4: Screen and filter the data information of the target case;

[0014] Step 5: Use the risk feature library model constructed in step 3 to analyze, calculate, extract and process the target case data information processed in step 4 to obtain the risk features and weights of the target cases;

[0015] Step 6: The front-end interface obtains and displays the target case data information processed in step 4, and displays the risk feature with the highest feature weight value in the target case and its related data.

[0016] Preferably, the screening and filtering of data information in step 1 is to use an unsupervised learning algorithm to identify abnormal information to obtain key information of the case instance, eliminate abnormal information, retain key information, and make the result more accurate.

[0017] Preferably, the risk feature extraction in step 2 uses a Random Forest algorithm to screen and extract features, and the use of the algorithm to extract features makes the result more reliable.

[0018] Preferably, the risk feature library model constructed in step 3 adopts a vector space model, which facilitates mutual calculation between cases.

[0019] Preferably, in step 5, after data information processing, feature extraction and weight calculation of the target case, various types of data of the target case are added to the risk feature library model each time, so as to update and iterate the model, which can continuously optimize the model.

[0020] Furthermore, the update iterative processing uses the Louvain community detection algorithm combined with the audit experiment to iteratively update and optimize the model, and the algorithm makes the model optimization and update more accurate.

[0021] Preferably, the data information displayed in step 6 will be presented in the form of images, charts and graphs according to different data information, which is more convenient and straightforward for users to watch.

[0022] Preferably, in step 6, data information, risk characteristics and related data are displayed, the association relationship between each data is displayed according to user needs, and relationship data and association paths at different levels are displayed, so that the data relationship display is more convenient and intuitive.

[0023] An intelligent risk prediction and identification device based on big data graph computing, characterized by comprising the following modules:

[0024] The filtering module is used to obtain multiple case instances and filter the data information in the cases;

[0025] Feature extraction module, used to extract risk features in each case instance;

[0026] Model building module, used to calculate the risk feature weights and the similarity of case instances, and build a risk feature library model;

[0027] The case processing module is used to use the model building module to analyze and calculate the target case and obtain the risk characteristics and weights of the target case;

[0028] The front-end display module is used to use graphics to display the correlation and correlation path between various data information and risk characteristics of the target case.

[0029] A smart risk prediction and identification device based on big data graph computing, characterized in that the device includes a service processor and a distributed memory, the service processor is connected to the memory, the distributed memory stores a service self-management program, and is configured to store machine-readable instructions, the service processor executes the service self-management program, and when the instructions are executed by the processor, the smart risk prediction and identification system based on big data graph computing as described above is realized.

[0030] The present invention analyzes and filters the data information of multiple historical case instances, extracts their risk features, calculates feature weights and similarities between cases, establishes a risk feature library model, uses the risk feature library model to analyze and calculate target cases, displays the obtained data information and risk features, and displays the correlation and paths between them, so that users can more conveniently view the data information and risk data they need.

[0031] 3. Beneficial effects

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] (1) The present invention obtains key data information by screening and filtering multiple case examples, removes abnormal information, extracts risk features using the obtained data information, and calculates feature weights and case similarities based on them to build a risk feature library model. The model established through past case data makes the calculation results more accurate;

[0034] (2) Each time the present invention uses the risk feature library model to analyze and calculate the target case, various data of the target case will be added to the risk feature library model, and the model will be updated and iterated. By updating and optimizing the model parameters and adding model data, the model after each calculation and processing will become more and more accurate and more in line with the development of the industry;

[0035] (3) The present invention uses images, charts and graphs to display different data information and their association relationships, and displays different levels of relationship data and association paths according to user needs, which is convenient for users to watch and query, and provides users with a clear viewing experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments or exemplary embodiments of the present application, the drawings required for use in the embodiments or exemplary descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application and should not be regarded as limiting the scope. For ordinary technicians in this field, other drawings can be obtained according to the drawings without paying creative work.

[0037] Figure 1 It is a schematic diagram of the steps of the present invention;

[0038] Figure 2 It is a schematic diagram of the process of the present invention;

[0039] Figure 3 It is a schematic diagram of the device of the present invention;

[0040] Figure 4 It is a schematic diagram of the device of the present invention;

[0041] Figure 5 It is a functional schematic diagram of the present invention. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0043] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0044] Example 1

[0045] This embodiment provides an intelligent risk prediction and identification system based on big data graph computing. The system is applied in the fields of big data analysis technology and knowledge graph technology. When there may be hidden risks in the case instances of an enterprise, but the enterprise users are unable to accurately obtain risk characteristics and related data, thereby threatening the development and operation of the enterprise, it is necessary to analyze and process the case instances of the enterprise to find out the hidden risk problems.

[0046] The intelligent risk prediction and identification system based on big data graph computing may include a historical case extraction system, a model database, a data processing device and a front-end display interface.

[0047] The historical case extraction system can extract information data from past cases and obtain key information from the cases.

[0048] The model database can extract the risk features of cases, calculate the similarity of cases, build a risk feature library model, and update the model based on new target cases. Some machine learning capabilities are implemented based on Spark MLlib.

[0049] The data processing device will use the model database to analyze and calculate the target case to obtain the risk characteristics of the target case.

[0050] The front-end display interface can display the data information required by the user and show the relationship between the data.

[0051] According to the above, the main process of the intelligent risk prediction and identification system based on big data graph calculation is as follows:

[0052] like Figure 2 As shown, firstly, multiple historical case instances are obtained and roughly classified, and at the same time, the data information in the historical case instances is screened and filtered, and an unsupervised learning algorithm is used to identify and process abnormal information to obtain key information of the case instances.

[0053] Then, based on the screening and filtering results, risk features are extracted for each historical case instance. The risk feature extraction uses the Random forest algorithm to screen and extract features, calculate the risk feature weight and the similarity of each case instance, and then use the vector space model to construct a risk feature library model based on the calculated risk feature weight and case instance similarity.

[0054] Obtain the target case, screen and filter the data information of the target case, use the constructed risk feature library model to analyze, calculate and extract the data information of the target case, obtain the risk characteristics of the target case and its weight, and after data information processing, feature extraction and weight calculation of the target case, each time add various types of data of the target case to the risk feature library model to update and iterate the model. The update and iterative processing uses the Louvain community detection algorithm combined with the audit experiment to iteratively update and optimize the model.

[0055] The front-end interface obtains and displays the processed target case data information, and displays the risk features with the highest feature weight values ​​and their related data in the target case. The displayed data information will be presented in the form of images, charts and graphs according to different data information. The association between each data will be displayed according to user needs, and the relationship data and association paths at different levels will be displayed.

[0056] From the above description, it can be seen that in this example, the data information of multiple historical case instances is analyzed and filtered to extract their risk characteristics, calculate the characteristic weights and the similarity between cases, establish a risk characteristic library model, and use the risk characteristic library model to analyze and calculate the target case. The obtained data information and risk characteristics are displayed, and the correlation and path between them are displayed. By continuously optimizing the risk characteristic library model, the model after each calculation and processing becomes more and more accurate, thereby improving the efficiency of risk discovery.

[0057] Example 2

[0058] like Figure 3 As shown, an intelligent risk prediction and identification device based on big data graph computing is characterized by comprising the following modules:

[0059] The filtering module is used to obtain multiple case instances and filter the data information in the cases;

[0060] Feature extraction module, used to extract risk features in each case instance;

[0061] Model building module, used to calculate the risk feature weights and the similarity of case instances, and build a risk feature library model;

[0062] The case processing module is used to use the model building module to analyze and calculate the target case and obtain the risk characteristics and weights of the target case;

[0063] The front-end display module is used to use graphics to display the correlation and correlation path between various data information and risk characteristics of the target case.

[0064] From the above description, it can be seen that in this example, the data information of multiple historical case instances is analyzed and filtered, their risk characteristics are extracted, the characteristic weights and the similarity between cases are calculated, and a risk characteristic library model is established. The risk characteristic library model is used to analyze and calculate the target case, and the obtained data information and risk characteristics are displayed. At the same time, the correlation and path between them are displayed. Through the data transportation between each module, the operation efficiency is improved to better conform to the development of the industry.

[0065] Example 3

[0066] like Figure 4 As shown, an intelligent risk prediction and identification device based on big data graph computing is characterized in that the device includes a service processor and a distributed memory, the service processor is connected to the memory, the distributed memory stores a service self-management program, and is configured to store machine-readable instructions, the service processor executes the service self-management program, and the instructions, when executed by the processor, realize the intelligent risk prediction and identification system based on big data graph computing as described in Example 1.

[0067] From the above description, it can be seen that in this example, the data information of multiple historical case instances is analyzed and filtered, their risk features are extracted, the feature weights and the similarity between cases are calculated, and a risk feature library model is established. The risk feature library model is used to analyze and calculate the target case, and the obtained data information and risk features are displayed, and the correlation and path between them are displayed at the same time, which is convenient for users to watch and query, giving users a clear viewing experience.

[0068] The above-mentioned embodiments only express the preferred implementation modes of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for those skilled in the art, several modifications, improvements and substitutions can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention.

Claims

1. An intelligent risk prediction and identification system based on big data graph computing, characterized in that: Use the following steps: Step 1: Obtain multiple case instances, perform rough classification, and filter the data information in the case instances; Step 2: Extract risk features for each case instance based on the processing results of step 1, and calculate the risk feature weight and the similarity of each case instance; Step 3: Construct a risk feature library model based on the risk feature weights and case instance similarities calculated in step 2; Step 4: Screen and filter the data information of the target case; Step 5: Use the risk feature library model constructed in step 3 to analyze, calculate, extract and process the target case data information processed in step 4 to obtain the risk features and weights of the target cases; Step 6: The front-end interface obtains and displays the target case data information processed in step 4, and displays the risk feature with the highest feature weight value in the target case and its related data; The data information displayed in step 6 may be displayed in the form of images, charts and graphs according to different data information; In step 6, data information, risk characteristics and related data are displayed, and the association relationship between various data is displayed according to user needs, and relationship data and association paths at different levels are displayed; The screening and filtering of data information in step 1 is to use an unsupervised learning algorithm to identify abnormal information and obtain key information of the case instance; The risk feature extraction in step 2 uses the Random Forest algorithm to screen and extract the features; In step 3, the risk feature library model is constructed by using a vector space model; In step 5, after data information processing, feature extraction and weight calculation of the target case, each type of data of the target case is added to the risk feature library model to update and iterate the model; The update iterative process uses the Louvain community detection algorithm combined with an audit experiment to iteratively update and optimize the model.

2. An intelligent risk prediction and identification device based on big data graph computing, characterized in that: The system according to claim 1 is used, and comprises the following modules: The filtering module is used to obtain multiple case instances and filter the data information in the cases; Feature extraction module, used to extract risk features in each case instance; Model building module, used to calculate the risk feature weights and the similarity of case instances, and build a risk feature library model; The case processing module is used to use the model building module to analyze and calculate the target case and obtain the risk characteristics and weights of the target case; The front-end display module is used to use graphics to display the correlation and correlation path between various data information and risk characteristics of the target case.

3. An intelligent risk prediction and identification device based on big data graph computing, characterized in that: The device includes a service processor and a distributed memory, the service processor is connected to the memory, the distributed memory stores a service self-management program, and is configured to store machine-readable instructions. The service processor executes the service self-management program, and when the instructions are executed by the processor, they implement the intelligent risk prediction and identification method based on big data graph computing as described in claim 1.

Citation Information

Patent Citations

  • Enterprise risk identification monitoring method, device and equipment and storage medium

    CN109492945A

  • Enterprise risk analysis method, apparatus, computer device and storage medium

    CN109523116A

  • Risk early warning method and device based on big data and computer equipment

    CN109829628A

  • Entity relationship graph display method and system

    CN111309824A

  • Risk identification method and device based on artificial intelligence, computer equipment and medium

    CN111950625A