Change risk prediction method and related device

By constructing a risk conduction map and shortest path algorithm, the risk coefficient is calculated, and the problem of low accuracy of change risk prediction methods in the banking system is solved, and the accuracy and comprehensiveness of risk prediction are improved.

CN119065874APending Publication Date: 2024-12-03BANK OF CHINA
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411112971.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

The existing change risk prediction methods are less accurate and flexible in the banking system, making it difficult to effectively reduce the operation and maintenance risks of data centers.

Method used

By obtaining the change attribute set of target change objects, a risk conduction map is constructed, a shortest path from the vertex of the risk conduction map to the target node is used to find the shortest path from the vertex of the risk conduction map to the target node, a risk coefficient is calculated, and the risk path is filtered according to preset conditions to obtain risk prediction data. To improve the accuracy and comprehensiveness of risk prediction.

Benefits of technology

It improves the accuracy and comprehensiveness of change risk prediction, can more accurately screen out the paths that are most affected by the target change object, and enhances the reliability of risk prediction results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119065874A_ABST
    Figure CN119065874A_ABST
Patent Text Reader

Abstract

The invention discloses a change risk prediction method and a related device, which can be applied to the financial field or other fields. Comprising the following steps: taking a weight value of an edge between two nodes in the risk conduction graph as a distance, searching by using a preset shortest path algorithm to obtain a shortest path from a vertex of the risk conduction graph to each target node, and taking the shortest path as a candidate path; and calculating a risk coefficient of the candidate path based on the number and the weight of the edges from the vertex of the candidate path to each path node. And taking the candidate path meeting a preset risk conduction condition as a risk path of the target change object, and obtaining risk prediction data of the target change object. The shortest path represents the path closest to the target node from the vertex, and the risk conduction condition includes that the risk coefficient is smaller than a risk coefficient threshold or the number of path nodes is larger than a node number threshold. Therefore, the risk path is the shortest path with a large risk coefficient or multiple influence nodes, so that the accuracy of the risk prediction result is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of big data technology, and in particular to a change risk prediction method and related devices. Background Art

[0002] The data center of the banking system contains a large amount of business data and has extremely high availability requirements. The good operation of various equipment in the banking system is an important prerequisite for ensuring the stable operation of the data center. Due to the large number of equipment types and complex structures, a change often causes a large change risk. Accurate risk prediction of changes can effectively reduce the operation and maintenance risks of the data center.

[0003] Existing change risk prediction methods often determine change risks according to change process specifications. The accuracy and flexibility of the prediction methods are low. Therefore, how to improve the prediction accuracy of change risks is a technical problem that needs to be solved urgently by technical personnel in this field. Summary of the invention

[0004] In view of the above problems, the present application provides a change risk prediction method and related devices to achieve the purpose of improving the prediction accuracy of change risk. The specific scheme is as follows:

[0005] The first aspect of the present application provides a change risk prediction method, comprising:

[0006] Acquire a change attribute set of a target change object, the change attribute set including at least one change attribute, and the target change object is a change object to be predicted;

[0007] Based on the set of changed attributes of the target change object, a risk transmission graph is constructed with the target change object as the starting point. The risk transmission graph is a weighted directed graph with objects as nodes and associations between objects as edges. The direction of the edge represents the risk transmission direction, and the weight of the edge represents the risk transmission strength, and the weight is inversely correlated with the risk transmission strength.

[0008] The weight of the edge between two nodes in the risk transmission graph is used as the distance between the two nodes, and a preset shortest path algorithm is used to find the shortest path from the vertex of the risk transmission graph to each target node as a candidate path; the vertex of the risk transmission graph is the node representing the target change object, and the target node includes other nodes on the risk transmission graph except the vertex;

[0009] Calculate the risk coefficient of the candidate path based on the number and weight of the edges from the vertices of the candidate path to each path node;

[0010] The candidate path that meets the preset risk transmission condition is used as the risk path of the target change object, and the risk transmission condition includes that the risk coefficient is less than a preset risk coefficient threshold or the number of path nodes is greater than a preset node number threshold;

[0011] The risk prediction data of the target change object is obtained, wherein the risk prediction data includes path data of each of the risk paths, and the path data includes a risk coefficient and an object represented by each path node.

[0012] In a possible implementation, obtaining a set of changed attributes of a target changed object includes:

[0013] Obtaining a description of the change item to be predicted;

[0014] Inputting the description of the change item to be predicted into a pre-trained semantic segmentation model, obtaining a set of change attributes of the change object output by the semantic segmentation model as the set of change attributes of the target change object;

[0015] The semantic segmentation model is trained based on historical change data, wherein the historical change data includes a description of changes of sample changes with labels, and the labels include a change object of the sample change and a corresponding set of change attributes.

[0016] In a possible implementation, based on the change attribute set of the target change object, a risk transmission map is constructed with the target change object as a starting point, including:

[0017] Obtain a global knowledge graph, where the global knowledge graph is a directed graph with objects as nodes and relationships between objects as edges, and the direction of the edges is from the previous order node to the next order node;

[0018] Obtaining a global path with a node representing the target change object in the global knowledge graph as a vertex;

[0019] For each of the global paths, along the direction of the edge in the global path, based on the influence metric values ​​between the target attribute of the predecessor node of the target edge and the respective attributes of the successor node of the target edge, the influence metric value of the target edge is calculated until the influence metric value of the target edge is less than a preset first metric threshold, and the predecessor node of the target edge is used as the termination node to obtain a local path of the global path;

[0020] Wherein, the target edge is any edge, the impact metric value is positively correlated with the impact degree, the target attribute of the predecessor node of the initial edge includes each change attribute in the change attribute set, the target attribute of the predecessor node of the non-initial edge includes an associated attribute, the associated attribute includes an attribute whose impact metric value with the predecessor target attribute is greater than a preset second metric threshold, and the predecessor target attribute includes the target attribute of the predecessor edge of the non-initial edge;

[0021] A subgraph including local paths of all global paths is extracted from the global knowledge graph, and the subgraph is labeled with the mapping value of the edge impact metric value to obtain the risk transmission graph, wherein the impact metric value is inversely correlated with the mapping value.

[0022] In a possible implementation, calculating the influence metric value of the target edge based on the influence metric values ​​between the target attribute of the preceding node of the target edge and each attribute of the succeeding node of the target edge includes:

[0023] Obtaining a risk matrix between each preceding node and each succeeding node, wherein the risk matrix is ​​used to represent an impact measurement value between each attribute of the preceding node and each attribute of the succeeding node;

[0024] For each target attribute of the preceding node of the target edge, based on the risk matrix of the preceding node and the succeeding node of the target edge, obtain the influence metric value between the target attribute and each attribute of the succeeding node of the target edge; perform weighted summation on each influence metric value to obtain the influence metric value of the target attribute on the succeeding node;

[0025] The influence metric values ​​of all target attributes of the preceding nodes of the target edge on the succeeding nodes are added together to obtain the influence metric value of the target edge.

[0026] In a possible implementation, the risk coefficient of the candidate path is calculated based on the number and weight of the edges from the vertices of the candidate path to each path node, including:

[0027] For each path node, the jump coefficient of the path node is determined according to the number of edges from the vertex to the path node. The jump coefficient is positively correlated with the number of edges.

[0028] Calculate the product of the jump coefficient of the path node and the weight of the path edge to obtain the node risk value of the path node. The path edge is the edge pointing to the path node.

[0029] The average of the node risk values ​​of all path nodes on the candidate path is used as the risk coefficient of the candidate path.

[0030] In a possible implementation, the shortest path algorithm includes: Dijkstra algorithm.

[0031] A second aspect of the present application provides a change risk prediction device, comprising:

[0032] A change object determination unit, configured to obtain a change attribute set of a target change object, wherein the change attribute set includes at least one change attribute, and the target change object is a change object to be predicted;

[0033] A risk transmission graph construction unit, configured to construct a risk transmission graph based on the set of change attributes of the target change object and taking the target change object as a starting point, wherein the risk transmission graph is a weighted directed graph with objects as nodes and associations between objects as edges, wherein the direction of the edge represents the risk transmission direction, the weight of the edge represents the risk transmission strength, and the weight is inversely correlated with the risk transmission strength;

[0034] A shortest path search unit, used to use the weight of the edge between two nodes in the risk transmission graph as the distance between the two nodes, and use a preset shortest path algorithm to find the shortest path from the vertex of the risk transmission graph to each target node as a candidate path; the vertex of the risk transmission graph is the node representing the target change object, and the target node includes other nodes on the risk transmission graph except the vertex;

[0035] A path risk calculation unit, used to calculate the risk coefficient of the candidate path based on the number and weight of the edges from the vertices of the candidate path to each path node;

[0036] A risk path prediction unit, used to use the candidate path that meets a preset risk transmission condition as the risk path of the target change object, wherein the risk transmission condition includes that the risk coefficient is less than a preset risk coefficient threshold or the number of path nodes is greater than a preset node number threshold;

[0037] The risk data acquisition unit is used to acquire the risk prediction data of the target change object, wherein the risk prediction data includes the path data of each risk path, and the path data includes the risk coefficient and the object represented by each path node.

[0038] The third aspect of the present application provides a computer program product, including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements the change risk prediction method of the above-mentioned first aspect or any implementation method of the first aspect.

[0039] A fourth aspect of the present application provides an electronic device, comprising at least one processor and a memory connected to the processor, wherein:

[0040] The memory is used to store computer programs;

[0041] The processor is used to execute the computer program so that the electronic device can implement the change risk prediction method of the above-mentioned first aspect or any implementation method of the first aspect.

[0042] A fifth aspect of the present application provides a computer storage medium, which carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement the change risk prediction method of the above-mentioned first aspect or any implementation method of the first aspect.

[0043] By means of the above technical scheme, the present application provides a change risk prediction method and related devices, which obtain a set of change attributes of a target change object, wherein the set of change attributes includes at least one change attribute, and the target change object is a change object to be predicted. Based on the set of change attributes of the target change object, a risk transmission graph is constructed with the target change object as the starting point, and the weight of the edge between two nodes in the risk transmission graph is used as the distance between the two nodes. The shortest path from the vertex of the risk transmission graph to each target node is found using a preset shortest path algorithm as a candidate path. The vertex of the risk transmission graph is a node representing the target change object, and the target node includes other nodes on the risk transmission graph except the vertex. Based on the number and weight of the edges from the vertex of the candidate path to each path node, the risk coefficient of the candidate path is calculated. The candidate path that meets the preset risk transmission conditions is used as the risk path of the target change object, and the risk prediction data of the target change object is obtained. The risk prediction data includes the path data of each risk path, and the path data includes the risk coefficient and the object represented by each path node. Since the risk transmission graph is a weighted directed graph with objects as nodes and the associations between objects as edges, the direction of the edge represents the risk transmission direction, the weight of the edge represents the risk transmission intensity, and the weight is inversely correlated with the risk transmission intensity. Therefore, the shortest path represents the path with the shortest distance from the vertex to the target node, that is, the path with the strongest risk transmission. In addition, since the risk transmission conditions include the risk coefficient being less than the preset risk coefficient threshold or the number of path nodes being greater than the preset node number threshold. It can be seen that the risk path is the shortest path with a large risk coefficient or a large number of affected nodes. Therefore, the present application first uses the shortest path algorithm to screen the path that is most affected by the target change object for each object, that is, the candidate path, and then calculates the risk coefficient based on each candidate path. Thus, the risk coefficient and the risk prediction data obtained by the risk object are obtained in units of paths, which improves the accuracy and comprehensiveness of the risk prediction results. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0045] Figure 1 A schematic diagram of a system architecture provided for an embodiment of the present application;

[0046] Figure 2 A schematic diagram of an optional hardware structure of a terminal provided in an embodiment of the present application;

[0047] Figure 3 A schematic diagram of an optional hardware structure of a server provided in an embodiment of the present application;

[0048] Figure 4 A flow chart of a change risk prediction method provided in an embodiment of the present application;

[0049] Figure 5 A schematic diagram of a specific implementation flow of a change risk prediction method provided in an embodiment of the present application;

[0050] Figure 6 A schematic diagram of the specific structure of a risk transmission diagram provided in an embodiment of the present application;

[0051] Figure 7 A schematic diagram of the structure of a change risk prediction device provided in an embodiment of the present application;

[0052] Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0053] The following describes the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. The terms used in the implementation method section of the present application are only used to explain the specific embodiments of the present application, and are not intended to limit the present application.

[0054] The embodiments of the present application are described below in conjunction with the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0055] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and need not be used to describe a specific order or sequential order. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances, which is only to describe the distinction mode adopted by the objects of the same attributes when describing in the embodiments of the present application. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, so that the process, method, system, product or equipment comprising a series of units need not be limited to those units, but may include other units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0056] The present application can be applied in the field of data technology. Taking the banking system as an example, the hardware and software connection structure of the equipment in the banking system is complex, and there are a large number of changes. Each change not only affects the change object (such as the first server) but also affects the objects related to the change object (for example, the second server that has a hardware connection with the first server). In addition, since the equipment has a large number of attributes, changes in different attributes affect different attributes of the associated objects. In order to improve the accuracy and comprehensiveness of the change risk prediction, the present application uses the change attributes of the change object as the fine-grainedness, analyzes and predicts the changes to obtain change risk prediction data, thereby improving the comprehensiveness and accuracy of the prediction results.

[0057] See also Figure 1 , Figure 1 A schematic diagram of a system architecture is shown. The system may include a terminal 100 and a server 200. The server 200 may include one or more servers ( Figure 1 In the example, a server is included, and the server 200 can provide the method provided in the embodiment of the present application for one or more terminals.

[0058] Among them, a change risk prediction application can be installed on the terminal 100. The above application and web page can provide an interface. The terminal 100 can receive relevant parameters entered by the user on the pre-configured interface and send the above parameters to the server 200. The server 200 can obtain processing results based on the received parameters and return the processing results to the terminal 100.

[0059] It should be understood that in some optional implementations, the terminal 100 can also complete the action of obtaining the processing result based on the received parameters by itself without the cooperation of the server, and the embodiments of the present application are not limited to this.

[0060] Next describe Figure 1 The product form of the mid-terminal 100;

[0061] The terminal 100 in the embodiment of the present application can be a mobile phone, a tablet computer, a wearable device, a vehicle-mounted device, an augmented reality device, or a wearable device. Augmented reality (AR) / virtual reality (VR) devices, laptops, ultra-mobile personal computers Human Computing Ultra-mobile personal computer (UMPC), netbook, personal digital assistant (PDA), etc., and the embodiments of the present application do not impose any limitation on this.

[0062] Figure 2 An optional hardware structure diagram of the terminal 100 is shown.

[0063] refer to Figure 2 As shown, the terminal 100 may include a radio frequency unit 110, a memory 120, an input unit 130, a display unit 140, a camera 150 (optional), an audio circuit 160 (optional), a speaker 161 (optional), a microphone 162 (optional), an earphone jack 163 (optional), a processor 170, an external interface 180, a power supply 190 and other components. Those skilled in the art will appreciate that Figure 2 These are merely examples of terminals or multi-function devices and do not constitute limitations on the terminals or multi-function devices, which may include more or fewer components than those shown in the figures, or combinations of certain components, or different components.

[0064] The input unit 130 can be used to receive input digital or character information, and generate key signal input related to the user settings and function control of the portable multifunctional device. Specifically, the input unit 130 may include a touch screen 131 (optional) or other input devices 132. The touch screen 131 can collect the user's touch operations on or near it (such as the user's operation on or near the touch screen using any suitable object such as fingers, joints, stylus, etc.), and drive the corresponding connection device according to a pre-set program. The touch screen can detect the user's touch action on the touch screen, convert the touch action into a touch signal and send it to the processor 170, and can receive and execute the command sent by the processor 170; the touch signal at least includes the touch point coordinate information. The touch screen 131 can provide an input interface and an output interface between the terminal 100 and the user. In addition, the touch screen can be implemented using multiple types such as resistive, capacitive, infrared and surface acoustic wave. In addition to the touch screen 131, the input unit 130 can also include other input devices. Specifically, other input devices 132 may include, but are not limited to, one or more of a physical keyboard, function keys (such as a volume control key, a switch key, etc.), a trackball, a mouse, a joystick, and the like.

[0065] Among them, the input device 132 can receive input data and the like.

[0066] The display unit 140 can be used to display information input by the user or provided to the user, various menus of the terminal 100, interactive interfaces, file display or playback of any multimedia file. In the embodiment of the present application, the display unit 140 can be used to display the interface of the data read and write instructions, processing results, etc.

[0067] The memory 120 can be used to store instructions and data. The memory 120 can mainly include an instruction storage area and a data storage area. The data storage area can store various data, such as multimedia files, texts, etc.; the instruction storage area can store software units such as operating systems, applications, instructions required for at least one function, or their subsets and extensions. It can also include a non-volatile random access memory; provide the processor 170 with hardware, software and data resources including management of computing and processing equipment, and support control software and applications. It is also used for the storage of multimedia files, and the storage of running programs and applications.

[0068] The processor 170 is the control center of the terminal 100. It uses various interfaces and lines to connect various parts of the entire terminal 100. By running or executing instructions stored in the memory 120 and calling data stored in the memory 120, it executes various functions of the terminal 100 and processes data, thereby controlling the terminal device as a whole. Optionally, the processor 170 may include one or more processing units; preferably, the processor 170 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface and application program, and the modem processor mainly processes wireless communication. It is understandable that the above-mentioned modem processor may not be integrated into the processor 170. In some embodiments, the processor and the memory may be implemented on a single chip, and in some embodiments, they may also be implemented separately on separate chips. The processor 170 may also be used to generate corresponding operation control signals, send them to corresponding components of the computing and processing device, read and process data in the software, especially read and process data and programs in the memory 120, so that each functional module therein performs corresponding functions, thereby controlling the corresponding components to act according to the requirements of the instructions.

[0069] Among them, the memory 120 can be used to store software code related to a change risk prediction method, the processor 170 can execute the steps of a change risk prediction method, and can also schedule other units (such as the above-mentioned input unit 130 and display unit 140) to achieve corresponding functions.

[0070] The radio frequency unit 110 (optional) can be used for receiving and sending information or receiving and sending signals during a call, for example, after receiving the downlink information of the base station, it is sent to the processor 170 for processing; in addition, the designed uplink data is sent to the base station. Generally, the RF circuit includes but is not limited to an antenna, at least one amplifier, a transceiver, a coupler, a low noise amplifier (Low Noise Amplifier, LNA), a duplexer, etc. In addition, the radio frequency unit 110 can also communicate with network devices and other devices through wireless communication. The wireless communication can use any communication standard or protocol, including but not limited to Global System of Mobile communication (Global System of Mobile communication, GSM), General Packet Radio Service (General Packet Radio Service, GPRS), Code Division Multiple Access (Code Division Multiple Access, CDMA), Wideband Code Division Multiple Access (Wideband Code Division Multiple Access, WCDMA), Long Term Evolution (Long Term Evolution, LTE), email, Short Messaging Service (SMS), etc.

[0071] In this embodiment of the present application, the RF unit 110 can send data to the server 200 and receive processing results sent by the server 200.

[0072] It should be understood that the radio frequency unit 110 is optional and can be replaced by other communication interfaces, such as a network port.

[0073] The terminal 100 also includes a power supply 190 (such as a battery) for supplying power to various components. Preferably, the power supply can be logically connected to the processor 170 through a power management system, so that the power management system can manage functions such as charging, discharging, and power consumption.

[0074] The terminal 100 further includes an external interface 180 , which may be a standard Micro USB interface or a multi-pin connector, and may be used to connect the terminal 100 to communicate with other devices, or to connect a charger to charge the terminal 100 .

[0075] Although not shown, the terminal 100 may also include a flashlight, a wireless fidelity (WiFi) module, a Bluetooth module, sensors with different functions, etc., which are not described in detail here. Some or all of the methods described below may be applied in the following embodiments. Figure 2 In the terminal 100 shown.

[0076] Next describe Figure 1 The product form of the server 200;

[0077] Figure 3 A structural diagram of a server 200 is provided, such as Figure 3 As shown, the server 200 includes a bus 201, a processor 202, a communication interface 203, and a memory 204. The processor 202, the memory 204, and the communication interface 203 communicate with each other via the bus 201.

[0078] The bus 201 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0079] The processor 202 may be any one or more of a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP).

[0080] The memory 204 may include a volatile memory, such as a random access memory (RAM). The memory 204 may also include a non-volatile memory, such as a read-only memory (ROM), a flash memory, a hard drive (HDD), or a solid state drive (SSD).

[0081] Among them, the memory 204 can be used to store software codes related to a shard indexing method based on a hash ring, the processor 202 can execute the steps of a shard indexing method based on a hash ring of the chip, and can also schedule other units to implement corresponding functions.

[0082] It should be understood that the above-mentioned terminal 100 and server 200 can be centralized or distributed devices, and the processors in the above-mentioned terminal 100 and server 200 (such as processor 170 and processor 202) can be hardware circuits (such as application specific integrated circuit (ASIC), field-programmable gate array (FPGA), general-purpose processor, digital signal processor (DSP), microprocessor or microcontroller, etc.), or a combination of these hardware circuits. For example, the processor can be a hardware system with an instruction execution function, such as a CPU, DSP, etc., or a hardware system without an instruction execution function, such as an ASIC, FPGA, etc., or a combination of the above-mentioned hardware systems without an instruction execution function and hardware systems with an instruction execution function.

[0083] See also Figure 4 , Figure 4 A flow chart of a change risk prediction method provided in an embodiment of the present application is shown in FIG. Figure 4 As shown, the method includes S401 to S406, as follows:

[0084] S401: Obtain a set of changed attributes of a target changed object.

[0085] In this embodiment, the change attribute set includes at least one change attribute, and the target change object is the change object to be predicted.

[0086] In this embodiment, an object includes multiple attributes. For example, a server corresponds to hard disk attributes, memory attributes, and graphics card attributes. A change to the object may involve at least one attribute. Based on the description of the item to be predicted to be changed, the involved change objects and the corresponding change attributes are extracted.

[0087] S402. Based on the change attribute set of the target change object, a risk transmission diagram is constructed with the target change object as the starting point.

[0088] In this embodiment, the risk transmission graph is a weighted directed graph with objects as nodes and the associations between objects as edges. The direction of each edge in the risk transmission graph represents the direction of risk transmission, and the weight of the edge represents the intensity of risk transmission. The weight is inversely correlated with the intensity of risk transmission, that is, the greater the weight, the smaller the intensity of risk transmission, and the smaller the weight, the closer the relationship between nodes, and the greater the intensity of risk transmission.

[0089] It should be noted that when a directed edge is used to represent an association relationship, the pre-order node of the edge points to the post-order node, indicating that a change in the object represented by the pre-order node affects the object represented by the post-order node.

[0090] S403. Using the weight of the edge between two nodes in the risk transmission graph as the distance between the two nodes, a preset shortest path algorithm is used to find the shortest path from the vertex of the risk transmission graph to each target node as a candidate path.

[0091] In this embodiment, the vertices of the risk transmission graph are nodes representing target change objects, and the target nodes include other nodes on the risk transmission graph except the vertices.

[0092] S404: Calculate the risk coefficient of the candidate path based on the number and weight of the edges from the vertices of the candidate path to each path node.

[0093] In this embodiment, the vertices of the candidate paths are all vertices of the risk transmission graph, that is, nodes representing the target change objects, and the path nodes refer to the nodes on the candidate paths other than the vertices. The number of edges from the vertices to the path nodes represents the number of hops from the vertices to the path nodes, and the weights represent the degree of association. It can be understood that the greater the number of hops of the edge pointing to the path node, the less the path node is affected by the vertex, and the greater the weight of the edge pointing to the path node, the less the path node is affected by the previous node.

[0094] Based on this, based on the number and weight of the edges from the vertices of the candidate path to each path node, the risk coefficient of the candidate path calculated comprehensively integrates the degree to which each path node is affected by the vertex change.

[0095] S405. The candidate path that meets the preset risk transmission conditions is used as the risk path of the target change object.

[0096] In this embodiment, the risk transmission condition includes that the risk coefficient is less than a preset risk coefficient threshold or the number of path nodes is greater than a preset node number threshold.

[0097] S406: Obtain risk prediction data of the target change object.

[0098] In this embodiment, the risk prediction data includes path data of each risk path, and the path data includes a risk coefficient and an object represented by each path node.

[0099] It can be seen from the above technical scheme that a change risk prediction method provided by the embodiment of the present application obtains a set of change attributes of the target change object, constructs a risk transmission graph based on the set of change attributes of the target change object, takes the target change object as the starting point, takes the weight of the edge between two nodes in the risk transmission graph as the distance between the two nodes, and uses the preset shortest path algorithm to find the shortest path from the vertex of the risk transmission graph to each target node as a candidate path. Based on the number and weight of the edges from the vertex of the candidate path to each path node, the risk coefficient of the candidate path is calculated. The candidate path that meets the preset risk transmission condition is used as the risk path of the target change object, and the risk prediction data of the target change object is obtained. Since the risk transmission graph is a weighted directed graph with objects as nodes and the association relationship between objects as edges, the direction of the edge represents the risk transmission direction, the weight of the edge represents the risk transmission strength, and the weight is inversely correlated with the risk transmission strength. Therefore, the shortest path represents the path with the shortest distance from the vertex to the target node, that is, the path with the strongest risk transmission, and since the risk transmission condition includes that the risk coefficient is less than the preset risk coefficient threshold or the number of path nodes is greater than the preset node number threshold. It can be seen that the risk path is the shortest path with a large risk coefficient or many affected nodes. Therefore, the present application first uses the shortest path algorithm to screen the path for each object that is most affected by the target change object, that is, the candidate path, and then calculates the risk coefficient based on each candidate path. Thus, the risk coefficient and risk object are obtained in units of paths. The risk prediction data improves the accuracy and comprehensiveness of the risk prediction results.

[0100] Reference Figure 5 , Figure 5 A schematic diagram of a specific implementation flow of a change risk prediction method provided in an embodiment of the present application is shown in FIG. Figure 5 As shown, a change risk prediction method provided by an embodiment of the present application may include S501 to S512, and these steps are described in detail below.

[0101] S501. Obtain a description of a change item to be predicted.

[0102] S502: Input the description of the change item to be predicted into the pre-trained semantic segmentation model, and obtain the change attribute set of the change object output by the semantic segmentation model as the change attribute set of the target change object.

[0103] In this embodiment, the semantic segmentation model is trained based on historical change data, where the historical change data includes a description of changes of sample changes with labels, where the labels include a change object of the sample change and a corresponding set of change attributes.

[0104] This application uses a pre-trained semantic segmentation model to extract the change objects and change attributes in the predicted changes, thereby improving the efficiency and accuracy of change analysis.

[0105] S503: Obtain a global knowledge graph.

[0106] In this embodiment, the global knowledge graph is a directed graph with objects as nodes and the association relationships between objects as edges, and the direction of the edges is from the preceding node to the succeeding node.

[0107] It should be noted that the global knowledge graph includes all objects and the relationships between objects. The preceding node points to the succeeding node, which means that when the object represented by the preceding node changes, it will affect the succeeding node.

[0108] S504: Obtain a global path with the node representing the target change object in the global knowledge graph as a vertex.

[0109] In this embodiment, a global path is a path starting from a target change node (a node representing a target change object) and extending along the direction of the edge until reaching an endpoint.

[0110] S505. For each global path, along the direction of the edge in the global path, based on the influence metric values ​​between the target attribute of the predecessor node of the target edge and the various attributes of the successive nodes of the target edge, the influence metric value of the target edge is calculated until the influence metric value of the target edge is less than a preset first metric threshold, and the predecessor node of the target edge is used as the termination node to obtain the local path of the global path.

[0111] In this embodiment, the target edge is any edge, and the influence metric value is positively correlated with the influence degree, that is, the larger the influence metric value, the greater the influence degree.

[0112] In this embodiment, an edge with a predecessor node as a vertex is an initial edge, the target attribute of the predecessor node of the initial edge includes each change attribute in the change attribute set, the target attribute of the predecessor node of the non-initial edge includes an associated attribute, the associated attribute includes an attribute whose influence metric value with the predecessor target attribute is greater than a preset second metric threshold, and the predecessor target attribute includes the target attribute of the predecessor edge of the non-initial edge.

[0113] Taking the global path including A, B, C, and D along the edge direction as an example, the initial edge is AB, the non-initial edges include BC, and CD, the target attribute NA of the predecessor node of the initial edge, that is, vertex A, includes each changed attribute in the changed attribute set of object A, the target attribute NB of the predecessor node B of the non-initial edge BC includes the attribute of object B whose impact metric value with the target attribute of vertex A is greater than the second metric threshold, and the target attribute NC of the predecessor node C of the non-initial edge CD includes the attribute of object C whose impact metric value with the target attribute NB of node B is greater than the second metric threshold. That is, the impact metric value of each edge on the global path is only related to the attributes of the predecessor node and the successor node affected by this change.

[0114] In this embodiment, the local path of the global path takes the termination node as the final node. Continuing with the above example, the influence metric value of the AB edge and the influence metric value of the BC edge are both greater than the first metric threshold, and the influence metric value of the CD edge is less than the first metric threshold. Therefore, the local path of the global path includes A, B and C.

[0115] In this embodiment, an optional method for calculating the impact metric value includes the following (1) to (4):

[0116] (1) Obtain the risk matrix between each preceding node and succeeding node.

[0117] In this embodiment, the risk matrix is ​​pre-constructed, and the risk matrix is ​​used to represent the influence measurement value between each attribute of the previous node and each attribute of the subsequent node. Taking the attribute dimension as n, the risk matrix of the first node and the second node is an n×n matrix, and the value of the coordinate xy in the matrix represents the influence measurement value of the attribute x of the first node on the attribute y of the second node.

[0118] (2) For each target attribute of the preceding node of the target edge, based on the risk matrix of the preceding node and the succeeding node of the target edge, the influence metric value between the target attribute and each attribute of the succeeding node of the target edge is obtained.

[0119] (3) Perform a weighted summation on each influence metric value to obtain the influence metric value of the target attribute on the subsequent nodes.

[0120] In this embodiment, when performing weighted summation on each influence metric value, the importance of each attribute of the subsequent node may be used as the weight, which is not limited in this embodiment.

[0121] (4) Add the influence metrics of all target attributes of the preceding nodes of the target edge on the succeeding nodes to obtain the influence metric of the target edge.

[0122] Taking the target edge as AB edge, the target attributes of A as x1 and x2, and the attributes of B including y1 and y2 as an example, from the risk matrix of objects A and B, obtain the impact measurement value of x1 on y1 and the impact measurement value of x1 on y2, and perform weighted addition to obtain the impact measurement value X1 of x1 on object B, obtain the impact measurement value of x2 on y1 and the impact measurement value of x2 on y2, and perform weighted addition to obtain the impact measurement value X2 of x2 on object B, calculate X1+X2, and obtain the impact measurement value of this change of object A on object B, that is, the impact measurement value of the AB edge.

[0123] S506. Extract a subgraph of local paths including all global paths from the global knowledge graph, and annotate the subgraph with the mapping value of the edge impact metric to obtain a risk transmission graph.

[0124] In this embodiment, the impact metric value is inversely correlated with the mapping value. Optionally, the impact metric value is normalized to a decimal between 0 and 1, and the mapping value is equal to the reciprocal of the impact metric value. For example, if the impact metric value is 0.2, the mapping value is equal to 2.

[0125] In this embodiment, the risk transmission graph is a subgraph of local paths including all global paths. The vertices of the risk transmission graph are nodes representing the target change objects, and the edges in the risk transmission graph are labeled with weights, which are equal to the mapping values. It can be understood that the greater the degree of impact and the greater the impact measurement value, the smaller the weight used as the label.

[0126] Figure 6 A schematic diagram of a risk transmission diagram provided in an embodiment of the present application is shown as follows: Figure 6 As shown, the risk transmission diagram includes nodes 0 to 6, where node 0 is the vertex, that is, node 0 represents the target change object. The edge is labeled with a weight to indicate the distance between the nodes. For example, the edge between node 0 and node 1 points from node 0 to node 1 (the direction is not shown), and the weight is equal to 6. The larger the weight, the greater the distance and the smaller the impact.

[0127] S507. Using the weight of the edge between two nodes in the risk transmission graph as the distance between the two nodes, a preset shortest path algorithm is used to find the shortest path from the vertex of the risk transmission graph to each target node as a candidate path.

[0128] In this implementation, the target node includes other nodes on the risk conduction graph except the vertex. Optionally, the shortest path algorithm includes Dijkstra algorithm.

[0129] It should be noted that the specific method of using the Dijkstra algorithm to find the shortest path from the vertex of the risk transmission graph to each target node can be found in the prior art.

[0130] Still Figure 6 Taking the risk transmission diagram shown as an example, the candidate paths include:

[0131] The shortest path from node 0 to node 1 is S1 = [0-1];

[0132] The shortest path from node 0 to node 2 S2 = [0-2];

[0133] The shortest path from node 0 to node 3 is S3 = [0-1-3];

[0134] The shortest path from node 0 to node 4 is S4 = [0-1-3-4];

[0135] The shortest path from node 0 to node 5 S5 = [0-1-3-5];

[0136] The shortest path from node 0 to node 6 is S6 = [0-1-3-4-6].

[0137] S508: For each path node, determine the jump coefficient of the path node according to the number of edges from the vertex to the path node.

[0138] In this embodiment, the jump coefficient of the path node is positively correlated with the number of edges from the vertex to the path node. Taking candidate path S6 as an example, the hop count of node 1 is 1, the hop count of node 3 is 2, the hop count of node 4 is 3, and the hop count of node 6 is 4. Optionally, the hop count is normalized by dividing it by 10 to obtain the jump coefficient, the hop count of node 1 is 0.1, the hop count of node 3 is 0.2, the hop count of node 4 is 0.3, and the hop count of node 6 is 0.4.

[0139] S509: Calculate the product of the jump coefficient of the path node and the weight of the path edge to obtain the node risk value of the path node.

[0140] In this embodiment, the path edge is an edge pointing to a path node. For example, the path edge whose path node in S3 is node 3 is edge [1-3]. The node risk value of node 3 is equal to the weight 5 of edge [1-3] multiplied by the jump coefficient 0.2 of node 3, that is, 5×0.2=1.

[0141] S510: Taking the average of the node risk values ​​of all path nodes on the candidate path as the risk coefficient of the candidate path.

[0142] In this embodiment, the average of the node risk values ​​of all path nodes on the candidate path is taken as the arithmetic mean, that is, the node risk values ​​of all path nodes are added together and then divided by the number of path nodes on the candidate path.

[0143] Still Figure 6 Taking the risk transmission diagram shown as an example, the risk factors of the candidate paths include:

[0144] The risk factor of S1 = (0.1 × 2) = 0.2;

[0145] The risk factor of S2 = (0.1 × 6) = 0.6;

[0146] The risk factor of S3 = (0.1 × 2 + 0.2 × 5) ÷ 2 = 0.6;

[0147] The risk factor of S4 = (0.1×2+0.2×5+0.3×10)÷3=1.4;

[0148] The risk factor of S5 = (0.1 × 2 + 0.2 × 5 + 0.3 × 15) ÷ 3 = 1.9;

[0149] The risk coefficient of S6 = (0.1×2+0.2×5+0.3×10+0.4×2)÷4=1.25.

[0150] S511. The candidate path that meets the preset risk transmission conditions is used as the risk path of the target change object.

[0151] In this embodiment, the risk transmission condition includes any one of the following: the risk coefficient is less than a preset risk coefficient threshold or the number of path nodes is greater than a preset node number threshold, that is, the candidate path with a risk coefficient less than the preset risk coefficient threshold and the candidate path with the number of path nodes greater than the node number threshold are both risk paths for the target change object.

[0152] Taking the risk coefficient threshold as 1.2 and the node number threshold as 3 as an example, the risk coefficients of S1 and S2 are both less than 1.2, and the number of path nodes of S6 is greater than 3, then the risk path of the target change object includes S1, S2 and S6.

[0153] S512: Obtain risk prediction data of the target change object.

[0154] In this embodiment, the risk prediction data includes path data of each risk path, and the path data includes a risk coefficient and an object represented by each path node.

[0155] Still Figure 6 Taking the risk transmission diagram shown as an example, the risk prediction data of the target change object includes: the risk coefficient of the risk path S1 and the various path nodes included in S1, the risk coefficient of the risk path S2 and the various path nodes included in S2, and the risk coefficient of the risk path S6 and the various path nodes included in S6.

[0156] It should be noted that after obtaining risk prediction data in units of risk paths, the risk level of the change to be predicted can be obtained based on the number of risk paths included in the risk prediction data, the risk coefficient of the risk path, and the number of occurrences of each path node. For example, the more risk paths there are, the greater the risk coefficient of the risk path, and the more path nodes whose number of occurrences is greater than the prediction number threshold, the greater the risk level of the change to be predicted. The specific risk synthesis method can be executed based on a pre-configured risk level rule, and this embodiment does not limit this.

[0157] It can be seen from the above technical scheme that the embodiment of the present application provides a change risk prediction method, the risk transmission graph is a weighted directed graph with objects as nodes and the association relationship between objects as edges, the direction of the edge represents the risk transmission direction, the weight of the edge represents the risk transmission intensity, and the weight is inversely correlated with the risk transmission intensity. Therefore, the shortest path represents the path with the shortest distance from the vertex to the target node, that is, the path with the strongest risk transmission. In addition, since the risk transmission conditions include the risk coefficient being less than the preset risk coefficient threshold or the number of path nodes being greater than the preset node number threshold. It can be seen that the risk path is the shortest path with a large risk coefficient or a large number of affected nodes. Therefore, the present application first uses the shortest path algorithm to screen the path that is most affected by the target change object for each object, that is, the candidate path, and then calculates the risk coefficient based on each candidate path. Thus, the risk coefficient and the risk prediction data obtained by the risk object are obtained in units of paths, which improves the accuracy and comprehensiveness of the risk prediction results.

[0158] Furthermore, S503 to S506 provide a specific method for constructing a risk transmission map based on a set of change attributes of a target change object and taking the target change object as a starting point, wherein the local paths affected by the target change object to be predicted are obtained from a pre-constructed global knowledge graph, and subgraphs of the global knowledge graph are extracted based on all local paths, and the subgraphs are labeled based on the weights of each edge, thereby improving the efficiency of constructing the risk transmission map. Furthermore, since the weights of each edge are related to the attributes affected by the change, the weights of the edges are targeted, and the weights represent the degree of impact of the change with high accuracy.

[0159] Furthermore, S508 to S510 provide a specific method for calculating the risk coefficient of a candidate path based on the number and weights of edges from the vertices of the candidate path to each path node, wherein the risk coefficient obtained by comprehensively considering the number of hops between each path node and the vertex and the weights of the edges is more accurate.

[0160] It should be noted that the change risk prediction method and related device provided by the present invention can be used in the data processing field or the financial field. The above is only an example and does not limit the application field of the change risk prediction method and related device provided by the present invention.

[0161] A change risk prediction method provided in an embodiment of the present application is introduced above, and a device for executing the above change risk prediction method will be introduced below.

[0162] See also Figure 7 , Figure 7 This is a schematic diagram of the structure of a change risk prediction device provided in an embodiment of the present application. Figure 7 As shown, the change risk prediction device 700 includes:

[0163] A change object determination unit 701 is used to obtain a change attribute set of a target change object, wherein the change attribute set includes at least one change attribute, and the target change object is a change object to be predicted;

[0164] A risk transmission graph construction unit 702 is used to construct a risk transmission graph based on the change attribute set of the target change object and taking the target change object as the starting point. The risk transmission graph is a weighted directed graph with objects as nodes and associations between objects as edges. The direction of the edge represents the risk transmission direction, and the weight of the edge represents the risk transmission strength, and the weight is inversely correlated with the risk transmission strength.

[0165] The shortest path search unit 703 is used to use the weight of the edge between two nodes in the risk transmission graph as the distance between the two nodes, and use a preset shortest path algorithm to find the shortest path from the vertex of the risk transmission graph to each target node as a candidate path; the vertex of the risk transmission graph is the node representing the target change object, and the target node includes other nodes on the risk transmission graph except the vertex;

[0166] A path risk calculation unit 704, configured to calculate a risk coefficient of the candidate path based on the number and weight of edges from the vertices of the candidate path to each path node;

[0167] The risk path prediction unit 705 is used to use the candidate path that meets the preset risk transmission condition as the risk path of the target change object, wherein the risk transmission condition includes that the risk coefficient is less than a preset risk coefficient threshold or the number of path nodes is greater than a preset node number threshold;

[0168] The risk data acquisition unit 706 is used to acquire risk prediction data of the target change object, wherein the risk prediction data includes path data of each risk path, and the path data includes a risk coefficient and an object represented by each path node.

[0169] In a possible implementation, when the change object determination unit is used to obtain the change attribute set of the target change object, it is specifically used to:

[0170] Obtaining a description of the change item to be predicted;

[0171] Inputting the description of the change item to be predicted into a pre-trained semantic segmentation model, obtaining a set of change attributes of the change object output by the semantic segmentation model as the set of change attributes of the target change object;

[0172] The semantic segmentation model is trained based on historical change data, wherein the historical change data includes a description of changes of sample changes with labels, and the labels include a change object of the sample change and a corresponding set of change attributes.

[0173] In a possible implementation, the risk transmission map construction unit is used to construct the risk transmission map based on the change attribute set of the target change object and taking the target change object as the starting point, specifically for:

[0174] Obtain a global knowledge graph, where the global knowledge graph is a directed graph with objects as nodes and relationships between objects as edges, and the direction of the edges is from the previous order node to the next order node;

[0175] Obtaining a global path with a node representing the target change object in the global knowledge graph as a vertex;

[0176] For each of the global paths, along the direction of the edge in the global path, based on the influence metric values ​​between the target attribute of the predecessor node of the target edge and the respective attributes of the successor node of the target edge, the influence metric value of the target edge is calculated until the influence metric value of the target edge is less than a preset first metric threshold, and the predecessor node of the target edge is used as the termination node to obtain a local path of the global path;

[0177] Wherein, the target edge is any edge, the impact metric value is positively correlated with the impact degree, the target attribute of the predecessor node of the initial edge includes each change attribute in the change attribute set, the target attribute of the predecessor node of the non-initial edge includes an associated attribute, the associated attribute includes an attribute whose impact metric value with the predecessor target attribute is greater than a preset second metric threshold, and the predecessor target attribute includes the target attribute of the predecessor edge of the non-initial edge;

[0178] A subgraph including local paths of all global paths is extracted from the global knowledge graph, and the subgraph is labeled with the mapping value of the edge impact metric value to obtain the risk transmission graph, wherein the impact metric value is inversely correlated with the mapping value.

[0179] In a possible implementation, the risk conduction graph construction unit is used to calculate the impact metric value of the target edge based on the impact metric value between the target attribute of the preceding node of the target edge and each attribute of the succeeding node of the target edge, specifically for:

[0180] Obtaining a risk matrix between each preceding node and each succeeding node, wherein the risk matrix is ​​used to represent an impact measurement value between each attribute of the preceding node and each attribute of the succeeding node;

[0181] For each target attribute of the preceding node of the target edge, based on the risk matrix of the preceding node and the succeeding node of the target edge, obtain the influence metric value between the target attribute and each attribute of the succeeding node of the target edge; perform weighted summation on each influence metric value to obtain the influence metric value of the target attribute on the succeeding node;

[0182] The influence metric values ​​of all target attributes of the preceding nodes of the target edge on the succeeding nodes are added together to obtain the influence metric value of the target edge.

[0183] In a possible implementation, the path risk calculation unit is used to calculate the risk coefficient of the candidate path based on the number and weight of the edges from the vertices of the candidate path to each path node, specifically for:

[0184] For each path node, the jump coefficient of the path node is determined according to the number of edges from the vertex to the path node. The jump coefficient is positively correlated with the number of edges.

[0185] Calculate the product of the jump coefficient of the path node and the weight of the path edge to obtain the node risk value of the path node. The path edge is the edge pointing to the path node.

[0186] The average of the node risk values ​​of all path nodes on the candidate path is used as the risk coefficient of the candidate path.

[0187] In a possible implementation, the shortest path algorithm includes: Dijkstra algorithm.

[0188] The present application also provides an electronic device in an embodiment. Figure 8 As shown, it shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present application. The electronic device in the embodiments of the present application may include but is not limited to fixed terminals such as mobile phones, notebook computers, PDAs (personal digital assistants), PADs (tablet computers), desktop computers, etc. Figure 8 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0189] like Figure 8As shown, the electronic device may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage device 808 to a random access memory (RAM) 803. When the electronic device is powered on, various programs and data required for the operation of the electronic device are also stored in the RAM 803. The processing device 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0190] Typically, the following devices may be connected to the I / O interface 805: input devices 806 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 807 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 808 including, for example, a memory card, a hard disk, etc.; and communication devices 809. The communication device 809 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although Figure 8 An electronic device having various devices is shown, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0191] An embodiment of the present application also provides a computer program product, including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements any one of the change risk prediction methods provided in the embodiments of the present application.

[0192] A computer-readable storage medium is also provided in an embodiment of the present application. The storage medium carries one or more computer programs. When one or more computer programs are executed by an electronic device, the electronic device can implement any change risk prediction method provided in an embodiment of the present application.

[0193] It should also be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the device embodiments provided by the present application, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines.

[0194] Through the description of the above implementation mode, the technicians in the field can clearly understand that the present application can be implemented by means of software plus necessary general hardware, and of course, it can also be implemented by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. In general, all functions completed by computer programs can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be various, such as analog circuits, digital circuits or special circuits. However, for the present application, software program implementation is a better implementation mode in more cases. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a readable storage medium, such as a computer floppy disk, a U disk, a mobile hard disk, a ROM, a RAM, a disk or an optical disk, etc., including a number of instructions to enable a computer device (which can be a personal computer, a training device, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0195] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0196] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a training device, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, training device, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device, a data center, etc. that includes one or more available media integrations. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

Claims

1. A change risk prediction method, characterized in that: include: Acquire a change attribute set of a target change object, the change attribute set including at least one change attribute, and the target change object is a change object to be predicted; Based on the set of changed attributes of the target change object, a risk transmission graph is constructed with the target change object as the starting point. The risk transmission graph is a weighted directed graph with objects as nodes and associations between objects as edges. The direction of the edge represents the risk transmission direction, and the weight of the edge represents the risk transmission strength, and the weight is inversely correlated with the risk transmission strength. The weight of the edge between two nodes in the risk transmission graph is used as the distance between the two nodes, and a preset shortest path algorithm is used to find the shortest path from the vertex of the risk transmission graph to each target node as a candidate path; the vertex of the risk transmission graph is the node representing the target change object, and the target node includes other nodes on the risk transmission graph except the vertex; Calculate the risk coefficient of the candidate path based on the number and weight of the edges from the vertices of the candidate path to each path node; The candidate path that meets the preset risk transmission condition is used as the risk path of the target change object, and the risk transmission condition includes that the risk coefficient is less than a preset risk coefficient threshold or the number of path nodes is greater than a preset node number threshold; The risk prediction data of the target change object is obtained, wherein the risk prediction data includes path data of each of the risk paths, and the path data includes a risk coefficient and an object represented by each path node.

2. The change risk prediction method according to claim 1, characterized in that: The acquisition of the change attribute set of the target change object includes: Obtaining a description of the change item to be predicted; Inputting the description of the change item to be predicted into a pre-trained semantic segmentation model, obtaining a set of change attributes of the change object output by the semantic segmentation model as the set of change attributes of the target change object; The semantic segmentation model is trained based on historical change data, wherein the historical change data includes a description of changes of sample changes with labels, and the labels include a change object of the sample change and a corresponding set of change attributes.

3. The change risk prediction method according to claim 1, characterized in that: The step of constructing a risk transmission diagram based on the set of changed attributes of the target change object and taking the target change object as a starting point includes: Obtain a global knowledge graph, where the global knowledge graph is a directed graph with objects as nodes and relationships between objects as edges, and the direction of the edges is from the previous order node to the next order node; Obtaining a global path with a node representing the target change object in the global knowledge graph as a vertex; For each of the global paths, along the direction of the edge in the global path, based on the influence metric values ​​between the target attribute of the predecessor node of the target edge and the respective attributes of the successor node of the target edge, the influence metric value of the target edge is calculated until the influence metric value of the target edge is less than a preset first metric threshold, and the predecessor node of the target edge is used as the termination node to obtain a local path of the global path; Wherein, the target edge is any edge, the impact metric value is positively correlated with the impact degree, the target attribute of the predecessor node of the initial edge includes each change attribute in the change attribute set, the target attribute of the predecessor node of the non-initial edge includes an associated attribute, the associated attribute includes an attribute whose impact metric value with the predecessor target attribute is greater than a preset second metric threshold, and the predecessor target attribute includes the target attribute of the predecessor edge of the non-initial edge; A subgraph including local paths of all global paths is extracted from the global knowledge graph, and the subgraph is labeled with the mapping value of the edge impact metric value to obtain the risk transmission graph, wherein the impact metric value is inversely correlated with the mapping value.

4. The change risk prediction method according to claim 3, characterized in that: The calculating the influence metric value of the target edge based on the influence metric value between the target attribute of the preceding node of the target edge and each attribute of the succeeding node of the target edge includes: Obtaining a risk matrix between each preceding node and each succeeding node, wherein the risk matrix is ​​used to represent an impact measurement value between each attribute of the preceding node and each attribute of the succeeding node; For each target attribute of the preceding node of the target edge, based on the risk matrix of the preceding node and the succeeding node of the target edge, obtain the influence metric value between the target attribute and each attribute of the succeeding node of the target edge; perform weighted summation on each influence metric value to obtain the influence metric value of the target attribute on the succeeding node; The influence metric values ​​of all target attributes of the preceding nodes of the target edge on the succeeding nodes are added together to obtain the influence metric value of the target edge.

5. The change risk prediction method according to claim 1, characterized in that: The step of calculating the risk coefficient of the candidate path based on the number and weight of the edges from the vertices of the candidate path to each path node includes: For each path node, the jump coefficient of the path node is determined according to the number of edges from the vertex to the path node. The jump coefficient is positively correlated with the number of edges. Calculate the product of the jump coefficient of the path node and the weight of the path edge to obtain the node risk value of the path node. The path edge is the edge pointing to the path node. The average of the node risk values ​​of all path nodes on the candidate path is used as the risk coefficient of the candidate path.

6. The change risk prediction method according to claim 1, characterized in that: The shortest path algorithm includes: Dijkstra algorithm.

7. A change risk prediction device, characterized in that: include: A change object determination unit, configured to obtain a change attribute set of a target change object, wherein the change attribute set includes at least one change attribute, and the target change object is a change object to be predicted; A risk transmission graph construction unit, configured to construct a risk transmission graph based on the set of change attributes of the target change object and taking the target change object as a starting point, wherein the risk transmission graph is a weighted directed graph with objects as nodes and associations between objects as edges, wherein the direction of the edge represents the risk transmission direction, the weight of the edge represents the risk transmission strength, and the weight is inversely correlated with the risk transmission strength; A shortest path search unit, used to use the weight of the edge between two nodes in the risk transmission graph as the distance between the two nodes, and use a preset shortest path algorithm to find the shortest path from the vertex of the risk transmission graph to each target node as a candidate path; the vertex of the risk transmission graph is the node representing the target change object, and the target node includes other nodes on the risk transmission graph except the vertex; A path risk calculation unit, used to calculate the risk coefficient of the candidate path based on the number and weight of the edges from the vertices of the candidate path to each path node; A risk path prediction unit, used to use the candidate path that meets a preset risk transmission condition as the risk path of the target change object, wherein the risk transmission condition includes that the risk coefficient is less than a preset risk coefficient threshold or the number of path nodes is greater than a preset node number threshold; The risk data acquisition unit is used to acquire the risk prediction data of the target change object, wherein the risk prediction data includes the path data of each risk path, and the path data includes the risk coefficient and the object represented by each path node.

8. A computer program product, characterized in that It includes computer-readable instructions, and when the computer-readable instructions are executed on an electronic device, the electronic device implements the change risk prediction method as described in any one of claims 1 to 6.

9. An electronic device, characterized in that: The method comprises at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer program so that the electronic device can implement the change risk prediction method as described in any one of claims 1 to 6.

10. A computer storage medium, characterized in that: The storage medium carries one or more computer programs, and when the one or more computer programs are executed by an electronic device, the electronic device can implement the change risk prediction method as described in any one of claims 1 to 6.

Citation Information

Cited By

  • Risk dynamic assessment method, system and terminal for non-standard lease contract

    CN120893821A

  • A risk dynamic evaluation method, system and terminal for non-standard lease contract

    CN120893821B

  • Cable professional knowledge graph construction method based on bright large model

    CN122529034A