CN112184012B - Enterprise risk early warning method, device, equipment and readable storage medium - Google Patents
Enterprise risk early warning method, device, equipment and readable storage medium Download PDFInfo
- Publication number
- CN112184012B CN112184012B CN202011036277.1A CN202011036277A CN112184012B CN 112184012 B CN112184012 B CN 112184012B CN 202011036277 A CN202011036277 A CN 202011036277A CN 112184012 B CN112184012 B CN 112184012B
- Authority
- CN
- China
- Prior art keywords
- node
- risk
- value
- enterprise
- association
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000011156 evaluation Methods 0.000 claims abstract description 41
- 238000012502 risk assessment Methods 0.000 claims abstract description 33
- 238000013528 artificial neural network Methods 0.000 claims description 10
- 230000007246 mechanism Effects 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 5
- 238000013210 evaluation model Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract description 3
- 238000010801 machine learning Methods 0.000 description 9
- 238000004422 calculation algorithm Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 230000008901 benefit Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000012954 risk control Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- General Business, Economics & Management (AREA)
- Computational Linguistics (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Animal Behavior & Ethology (AREA)
- Educational Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an enterprise risk early warning method, device and equipment and a readable storage medium, wherein the method comprises the following steps: receiving an early warning instruction; wherein, the early warning instruction includes: information of a plurality of entity objects and association relation information among the entity objects; drawing an enterprise association map according to the information of the plurality of entity objects and the association relation information among the entity objects; wherein, the enterprise association graph includes: an edge characterizing a relationship between a node of an entity object and the entity object; calculating a risk assessment value of each entity object by using a preset risk assessment model based on the enterprise association graph; judging whether the risk evaluation value of the entity object is larger than a preset threshold value, if so, sending a risk early warning message to the appointed terminal; the invention can more accurately perform enterprise risk early warning; the invention is applicable to the field of financial science and technology, and simultaneously relates to a blockchain technology.
Description
Technical Field
The invention relates to the technical field of intelligent decision making, in particular to an enterprise risk early warning method, an enterprise risk early warning device, enterprise risk early warning equipment and a readable storage medium.
Background
In recent years, domestic enterprise debt starts to break the law of just exchanging, the default event occurs, and the default quantity shows a growing trend. Therefore, risk control of bond markets and early warning of issue debts subjects at credit risk, avoiding significant losses for investors, are becoming increasingly important. The existing mode for judging whether the enterprise has the default risk is that the wind control personnel uses expert experience and reasoning logic thereof to judge the risk based on various bulletin information of the enterprise, and due to the limitations of self experience and visual angle coverage of different professionals, all cases are difficult to cover, and omission is easy to generate; with the continuous development of machine learning algorithms, machine learning algorithms are increasingly applied to various fields to solve, for example: data prediction, data classification and data clustering; however, the machine learning algorithm is applied to the field of enterprise default risk judgment, and the following technical problems exist: 1) How to process various bulletin information of an enterprise into information usable by a machine learning algorithm, and 2) how to improve the efficiency of the machine learning algorithm in utilizing various bulletin information of an enterprise.
Disclosure of Invention
The invention aims to provide an enterprise risk early warning method, device, equipment and storage medium, which can determine the risk influence relationship among enterprises, so that enterprise risk early warning can be more accurately carried out.
According to one aspect of the present invention, there is provided an enterprise risk early warning method, the method comprising:
Receiving an early warning instruction; wherein, the early warning instruction includes: information of a plurality of entity objects and association relation information among the entity objects;
Drawing an enterprise association map according to the information of the plurality of entity objects and the association relation information among the entity objects; wherein, the enterprise association graph includes: an edge characterizing a relationship between a node of an entity object and the entity object;
calculating a risk assessment value of each entity object by using a preset risk assessment model based on the enterprise association graph;
judging whether the risk evaluation value of the entity object is larger than a preset threshold value, if so, sending a risk early warning message to a designated terminal corresponding to the entity object.
Optionally, the drawing the enterprise association map according to the information of the plurality of entity objects and the association relationship information between the entity objects includes:
drawing nodes representing the respective entity objects, and setting node attributes for the nodes of each entity object; wherein the node attributes include: a preset potential risk value;
Drawing edges among the nodes according to the association relation information among the entity objects, and setting edge attributes for each edge to form an enterprise association graph; wherein the edge attributes include: the association vector is used for representing the association relation type, the association vector is a vector with 1 XN dimensions, and N is the category number of the association relation type.
Optionally, the calculating, based on the enterprise association graph and using a preset risk assessment model, a risk assessment value of each entity object includes:
aiming at a target node, determining a related node connected with the target node through an edge in the enterprise related graph;
Respectively calculating a risk evaluation value of each associated node, and respectively calculating an attention weight value of each associated node to the target node;
And calculating the risk evaluation value of the target node according to the risk evaluation value of each associated node, the attention weight value of each associated node to the target node and the potential risk value in the node attribute of the target node.
Optionally, the calculating the attention weight value of each associated node to the target node includes:
The attention weight value W ba of the associated node b to the target node a is calculated according to the following formula:
Wba＝softmax(Qa×Rab)；
Wherein Q a is the potential risk value in the node attribute of target node a;
R ab is the association vector in the edge attribute of the edge between the target node a and the association node b.
Optionally, the calculating the risk assessment value of the target node according to the risk assessment value of each associated node, the attention weight value of each associated node to the target node, and the potential risk value in the node attribute of the target node includes:
The conduction risk value V a- of the associated node to the target node a is calculated according to the following formula:
Va-＝f(Wba×Vb+Wca×Vc+…)；
wherein f () is a sigmoid function;
V b is a risk evaluation value of the associated node b, V c is a risk evaluation value of the associated node c;
w ba is the attention weight value of the associated node b to the target node a, and W ca is the attention weight value of the associated node c to the target node a;
the risk assessment value V a+ of the target node a is calculated according to the following formula:
Va+＝Va-+Va；
Where V a is a potential risk value in the node attribute according to target node a.
Optionally, the risk assessment model is a graph-annotation-force mechanism neural network GAT model.
In order to achieve the above object, the present invention further provides an enterprise risk early warning device, the device comprising:
The receiving module is used for receiving the early warning instruction; wherein, the early warning instruction includes: information of a plurality of entity objects and association relation information among the entity objects;
The drawing module is used for drawing an enterprise association graph according to the information of the plurality of entity objects and the association relation information among the entity objects so as to form the enterprise association graph; wherein, the enterprise association graph includes: an edge characterizing a relationship between a node of an entity object and the entity object;
The computing module is used for computing a risk evaluation value of each entity object by utilizing a preset risk evaluation model based on the enterprise association graph;
And the early warning module is used for judging whether the risk evaluation value of the entity object is larger than a preset threshold value, and if so, sending a risk early warning message to a designated terminal corresponding to the entity object.
Optionally, the drawing module is configured to:
drawing nodes representing the respective entity objects, and setting node attributes for the nodes of each entity object; wherein the node attributes include: a preset potential risk value;
Drawing edges among the nodes according to the association relation information among the entity objects, and setting edge attributes for each edge to form an enterprise association graph; wherein the edge attributes include: the association vector is used for representing the association relation type, the association vector is a vector with 1 XN dimensions, and N is the category number of the association relation type.
In order to achieve the above object, the present invention further provides a computer device, which specifically includes: the system comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the steps of the enterprise risk early warning method when executing the computer program.
In order to achieve the above object, the present invention further provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the enterprise risk early warning method described above.
According to the enterprise risk early warning method, the enterprise risk early warning device, the enterprise risk early warning equipment and the enterprise risk early warning storage medium, the advertising information of each enterprise and the association relation between the enterprises are converted into the enterprise association map, and the information used for being input into the machine learning model is extracted from the enterprise association map, so that whether the enterprise has default risks or not is judged by the machine learning model; in the invention, the influence of the associated enterprises on the target enterprises is comprehensively considered instead of the self information of the target enterprises, so that whether the target enterprises have default risks or not is more comprehensively and accurately judged. In addition, in the invention, the bulletin information of each enterprise and the association relation between the enterprises are converted into the enterprise association map, so that the technical problem that the machine learning model cannot utilize the bulletin information of the enterprises is solved, and meanwhile, the risk assessment efficiency of the machine learning model on the enterprises is improved.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to designate like parts throughout the figures. In the drawings:
FIG. 1 is a schematic flow chart of an alternative enterprise risk early warning method according to the first embodiment;
Fig. 2 is a schematic diagram of an alternative composition structure of an enterprise risk early warning device according to a second embodiment;
fig. 3 is a schematic diagram of an alternative hardware architecture of a computer device according to the third embodiment.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
The embodiment of the invention provides an enterprise risk early warning method, as shown in fig. 1, which specifically comprises the following steps:
step S101: receiving an early warning instruction; wherein, the early warning instruction includes: information of a plurality of entity objects and association relationship information between the respective entity objects.
Specifically, the entity object may be: companies, institutions, natural people; the association relationship includes the following types: controlling, mortgage, guarantee, actual controlling person, legal person and business party.
The information of the plurality of entity objects and the association relation information among the entity objects are notice information stored in the blockchain node; uploading the information of the plurality of entity objects and the association relation information among the entity objects to the blockchain can ensure the security and the fair transparency to the user.
The blockchain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, encryption algorithm and the like. The blockchain (Blockchain), essentially a de-centralized database, is a string of data blocks that are generated in association using cryptographic methods, each of which contains information from a batch of network transactions for verifying the validity (anti-counterfeit) of its information and generating the next block. The blockchain may include a blockchain underlying platform, a platform product services layer, an application services layer, and the like.
Step S102: drawing an enterprise association map according to the information of the plurality of entity objects and the association relation information among the entity objects; wherein, the enterprise association graph includes: the nodes representing the entity objects and the edges representing the association between the entity objects.
Specifically, step S102 includes:
step A1: drawing nodes representing the respective entity objects, and setting node attributes for the nodes of each entity object; wherein the node attributes include: a preset potential risk value;
Step A2: drawing edges among the nodes according to the association relation information among the entity objects, and setting edge attributes for each edge to form an enterprise association graph; wherein the edge attributes include: the association vector is used for representing the association relation type, the association vector is a vector with 1 XN dimensions, and N is the category number of the association relation type.
In this embodiment, the enterprise association graph may be effectively represented by the enterprise information, the association party information, and the specific association relationship between the enterprise information and the association party information; the specific process for constructing the enterprise association graph is as follows: based on the enterprise information extracted from various notices and the relationship data among enterprises, important entity objects such as companies, institutions and natural people are abstracted into nodes in an enterprise association graph, corresponding state information of the important entity objects is expressed in the form of node attributes, various association relations (such as control, mortgage, guarantee and the like) among the companies, the companies and the institutions and between the companies and personnel are abstracted into edges between the nodes, and the state information of specific association relations is expressed in the form of edge attributes.
It should be noted that, the expert sets corresponding potential risk values for the entity objects in advance according to the basic information, the management information, the loan information and the like of each entity object; the potential risk value is used for representing the current default risk probability of the entity object; the greater the potential risk value of an entity object, the greater the probability that the entity object has a default risk. In addition, since there may be one or more types of association relationships between the respective entity objects, in order to be able to represent the various types of association relationships between the respective entity objects by one enterprise association graph, an association vector is employed in the present embodiment; the value of each dimension of the initial association vector is 0, and each dimension corresponds to one type of association relationship. If there are two types of association relations of the indirection and the business transaction party between the two entity objects, the value of the dimension representing the indirection and the business transaction party is set to be 1 in the association vector.
In this embodiment, the innovative association vector in vector form represents the relationship edge, and the form of the polygonal and directed graph is effectively converted into the form of a single side, so that the constructed enterprise association graph is more suitable for a GAT (Graph Attention Network, graph annotation force mechanism neural network) learning model framework based on a single side structure, and the complex GAT learning model framework based on the polygonal directed graph structure is avoided from being specially designed, so that the whole GAT learning model framework is simplified.
Step S103: and calculating the risk assessment value of each entity object by using a preset risk assessment model based on the enterprise association graph.
Specifically, the risk assessment model is a graph-annotation-force mechanism neural network GAT model; if the risk assessment value of an entity object is larger, the probability that the entity object will generate credit risk event in a future period is larger.
The GAT is a graph neural network model (GNN) based on a graph attention mechanism, has the end-to-end learning characteristic of a neural network, and reduces complex conduction path factors to mine the characteristic engineering work. The graph-annotation force mechanism neural network GAT model comprises: a plurality of GAT layers and a full link layer, and including in each GAT layer: an attention layer and a residual network layer. The node attribute of each node and the edge attribute of each edge are input into the graph meaning mechanical neural network GAT model, and the output of the graph meaning mechanical neural network GAT model is a risk evaluation value of each node. In addition, the graph-annotation-force mechanism neural network GAT model is a model trained in advance according to a historical sample set.
Further, step S103 includes:
Step B1: aiming at a target node, determining a related node connected with the target node through an edge in the enterprise related graph;
step B2: respectively calculating a risk evaluation value of each associated node, and respectively calculating an attention weight value of each associated node to the target node;
Step B3: and calculating the risk evaluation value of the target node according to the risk evaluation value of each associated node, the attention weight value of each associated node to the target node and the potential risk value in the node attribute of the target node.
It should be noted that, the risk evaluation value of each associated node is calculated in the manner of the above steps B1 to B3, that is, one associated node is used as the target node to perform the above steps B1 to B3, so as to obtain the risk evaluation value of the associated node. And if one target node does not have an associated node, taking the potential risk value of the target node as a risk evaluation value of the target node.
Still further, step B2 includes:
The attention weight value W ba of the associated node b to the target node a is calculated according to the following formula:
Wba＝softmax(Qa×Rab)；
Wherein Q a is the potential risk value in the node attribute of target node a;
R ab is the association vector in the edge attribute of the edge between the target node a and the association node b.
Still further, step B3 includes:
Step B31: the conduction risk value V a- of the associated node to the target node a is calculated according to the following formula:
Va-＝f(Wba×Vb+Wca×Vc+…)；
wherein f () is a sigmoid function;
V b is a risk evaluation value of the associated node b, V c is a risk evaluation value of the associated node c;
w ba is the attention weight value of the associated node b to the target node a, and W ca is the attention weight value of the associated node c to the target node a;
Step B32: the risk assessment value V a+ of the target node a is calculated according to the following formula:
Va+＝Va-+Va；
Where V a is a potential risk value in the node attribute according to target node a.
The GAT automatically learns the influence degree of the association node on the target node through an attention mechanism, namely, the importance of different association relations corresponding to each enterprise can be learned. And in the learning process, the state information of the enterprise is a very important characteristic factor. I.e. the comprehensive interaction of the enterprise's own state and the associated party is learned at the same time. In reality, the probability of risk conduction is not only related to the related party, but also is highly related to the self risk resistance, so that the GAT has the advantage of more accuracy compared with a learning algorithm which only pays attention to the importance of the relationship.
In addition, the risk assessment value of each entity object can be uploaded into the blockchain to prevent the risk assessment value from being tampered; other users may download risk assessment values for each physical object from the blockchain.
Step S104: judging whether the risk evaluation value of the entity object is larger than a preset threshold value, if so, sending a risk early warning message to a designated terminal corresponding to the entity object.
Because the probability of risk conduction among enterprises is higher with the relation abstraction degree of enterprise association relation, the relation types among enterprises are more, the difference among different relations is larger, the characteristics of combination enhancement effect and the like exist among different relations, the generalization performance of the existing model is poorer, and good effects cannot be obtained. In this embodiment, a new correlation enterprise credit risk conduction model based on GAT is provided, which can effectively and automatically learn different attention weights of different relationship types, and at the same time learn the importance of the relationship, also learn the influence factors of the state of the enterprise, that is, learn the comprehensive characteristics of the correlation party and the enterprise at the same time, and more comprehensively simulate the display condition. In addition, the embodiment also adopts a vector form to express the relation edges, effectively converts the form of the multi-edge and directed graph into a single-edge form, avoids specially designing a complex GAT structure based on the multi-edge directed graph, and enables the GAT structure based on single-edge learning to be applicable through simple modification. Compared with wind control service personnel, the model is automatically operated based on the full data set, so that global coverage can be effectively realized, and the efficiency is improved.
Example two
The embodiment of the invention provides an enterprise risk early warning device, as shown in fig. 2, which specifically comprises the following components:
a receiving module 201, configured to receive an early warning instruction; wherein, the early warning instruction includes: information of a plurality of entity objects and association relation information among the entity objects;
The drawing module 202 is configured to draw an enterprise association graph according to the information of the plurality of entity objects and the association relationship information between the entity objects; wherein, the enterprise association graph includes: an edge characterizing a relationship between a node of an entity object and the entity object;
the calculating module 203 is configured to calculate a risk assessment value of each entity object by using a preset risk assessment model based on the enterprise association graph;
and the early warning module 204 is configured to determine whether the risk assessment value of the entity object is greater than a preset threshold, and if yes, send a risk early warning message to a specified terminal corresponding to the entity object.
Specifically, the drawing module 202 is configured to:
drawing nodes representing the respective entity objects, and setting node attributes for the nodes of each entity object; wherein the node attributes include: a preset potential risk value;
Drawing edges among the nodes according to the association relation information among the entity objects, and setting edge attributes for each edge to form an enterprise association graph; wherein the edge attributes include: the association vector is used for representing the association relation type, the association vector is a vector with 1 XN dimensions, and N is the category number of the association relation type.
Further, the computing module 203 specifically includes:
The determining unit is used for determining the associated node connected with the target node through the edge in the enterprise associated graph aiming at the target node;
a calculation unit, configured to calculate a risk evaluation value of each associated node, and calculate an attention weight value of each associated node to the target node;
And the processing unit is used for calculating the risk evaluation value of the target node according to the risk evaluation value of each associated node, the attention weight value of each associated node to the target node and the potential risk value in the node attribute of the target node.
Still further, the computing unit is specifically configured to:
The attention weight value W ba of the associated node b to the target node a is calculated according to the following formula:
Wba＝softmax(Qa×Rab)；
Wherein Q a is the potential risk value in the node attribute of target node a;
R ab is the association vector in the edge attribute of the edge between the target node a and the association node b.
Still further, the processing unit is configured to:
The conduction risk value V a- of the associated node to the target node a is calculated according to the following formula:
Va-＝f(Wba×Vb+Wca×Vc+…)；
wherein f () is a sigmoid function;
V b is a risk evaluation value of the associated node b, V c is a risk evaluation value of the associated node c;
w ba is the attention weight value of the associated node b to the target node a, and W ca is the attention weight value of the associated node c to the target node a;
the risk assessment value V a+ of the target node a is calculated according to the following formula:
Va+＝Va-+Va；
Where V a is a potential risk value in the node attribute according to target node a.
Further, the risk assessment model is a graph-annotation-force mechanism neural network GAT model.
Example III
The present embodiment also provides a computer device, such as a smart phone, a tablet computer, a notebook computer, a desktop computer, a rack-mounted server, a blade server, a tower server, or a rack-mounted server (including an independent server or a server cluster formed by a plurality of servers) that can execute a program. As shown in fig. 3, the computer device 30 of the present embodiment includes at least, but is not limited to: a memory 301, a processor 302, which may be communicatively connected to each other via a system bus. It is noted that FIG. 3 only shows a computer device 30 having components 301-302, but it should be understood that not all of the illustrated components are required to be implemented, and that more or fewer components may alternatively be implemented.
In this embodiment, the memory 301 (i.e., readable storage medium) includes flash memory, a hard disk, a multimedia card, a card memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, and the like. In some embodiments, the memory 301 may be an internal storage unit of the computer device 30, such as a hard disk or memory of the computer device 30. In other embodiments, the memory 301 may also be an external storage device of the computer device 30, such as a plug-in hard disk provided on the computer device 30, a smart memory card (SMART MEDIA CARD, SMC), a Secure Digital (SD) card, a flash memory card (FLASH CARD), or the like. Of course, the memory 301 may also include both internal storage units of the computer device 30 and external storage devices. In this embodiment, the memory 301 is typically used to store an operating system and various types of application software installed on the computer device 30. In addition, the memory 301 can also be used to temporarily store various types of data that have been output or are to be output.
Processor 302 may be a central processing unit (Central Processing Unit, CPU), controller, microcontroller, microprocessor, or other data processing chip in some embodiments. The processor 302 is generally used to control the overall operation of the computer device 30.
Specifically, in this embodiment, the processor 302 is configured to execute a program of an enterprise risk early warning method stored in the processor 302, where the program of the enterprise risk early warning method is executed to implement the following steps:
Receiving an early warning instruction; wherein, the early warning instruction includes: information of a plurality of entity objects and association relation information among the entity objects;
Drawing an enterprise association map according to the information of the plurality of entity objects and the association relation information among the entity objects; wherein, the enterprise association graph includes: an edge characterizing a relationship between a node of an entity object and the entity object;
calculating a risk assessment value of each entity object by using a preset risk assessment model based on the enterprise association graph;
judging whether the risk evaluation value of the entity object is larger than a preset threshold value, if so, sending a risk early warning message to a designated terminal corresponding to the entity object.
The specific embodiment of the above method steps may refer to the first embodiment, and this embodiment is not repeated here.
Example IV
The present embodiment also provides a computer readable storage medium, such as a flash memory, a hard disk, a multimedia card, a card memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, a server, an App application store, etc., having stored thereon a computer program that when executed by a processor performs the following method steps:
Receiving an early warning instruction; wherein, the early warning instruction includes: information of a plurality of entity objects and association relation information among the entity objects;
Drawing an enterprise association map according to the information of the plurality of entity objects and the association relation information among the entity objects; wherein, the enterprise association graph includes: an edge characterizing a relationship between a node of an entity object and the entity object;
calculating a risk assessment value of each entity object by using a preset risk assessment model based on the enterprise association graph;
judging whether the risk evaluation value of the entity object is larger than a preset threshold value, if so, sending a risk early warning message to a designated terminal corresponding to the entity object.
The specific embodiment of the above method steps may refer to the first embodiment, and this embodiment is not repeated here.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.
Claims (5)
1. An enterprise risk early warning method, which is characterized by comprising the following steps:
Receiving an early warning instruction; wherein, the early warning instruction includes: information of a plurality of entity objects and association relation information among the entity objects;
Drawing an enterprise association map according to the information of the plurality of entity objects and the association relation information among the entity objects; wherein, the enterprise association graph includes: an edge characterizing a relationship between a node of an entity object and the entity object;
calculating a risk assessment value of each entity object by using a preset risk assessment model based on the enterprise association graph;
Judging whether the risk evaluation value of the entity object is larger than a preset threshold value, if so, sending a risk early warning message to a designated terminal corresponding to the entity object;
The drawing the enterprise association map according to the information of the plurality of entity objects and the association relation information among the entity objects comprises the following steps:
drawing nodes representing the respective entity objects, and setting node attributes for the nodes of each entity object; wherein the node attributes include: a preset potential risk value;
Drawing edges among the nodes according to the association relation information among the entity objects, and setting edge attributes for each edge to form an enterprise association graph; wherein the edge attributes include: the association vector is used for representing the association relation type, the association vector is a vector with 1 XN dimension, and N is the category number of the association relation type;
Based on the enterprise association graph, calculating a risk assessment value of each entity object by using a preset risk assessment model, including:
aiming at a target node, determining a related node connected with the target node through an edge in the enterprise related graph;
Respectively calculating a risk evaluation value of each associated node, and respectively calculating an attention weight value of each associated node to the target node, wherein the attention weight value is specifically as follows: calculating the attention weight W ba of the associated node b to the target node a according to W ba＝softmax(Qa×Rab), wherein Q a is a potential risk value in the node attribute of the target node a, and R ab is an associated vector in the edge attribute of the edge between the target node a and the associated node b;
according to the risk evaluation value of each associated node, the attention weight value of each associated node to the target node and the potential risk value in the node attribute of the target node, the risk evaluation value of the target node is calculated, specifically: according to V a-＝f(Wba×Vb+Wca×Vc + …), calculating a conduction risk value V a- of the associated node to the target node a, wherein f () is a sigmoid function, V b is a risk evaluation value of the associated node b, V c is a risk evaluation value of the associated node c, W ba is an attention weight value of the associated node b to the target node a, W ca is an attention weight value of the associated node c to the target node a, and calculating a risk evaluation value V a+ of the target node a according to V a+＝Va-+Va, wherein V a is a potential risk value in a node attribute according to the target node a.
2. The enterprise risk early warning method of claim 1, wherein the risk assessment model is a graph-effort mechanism neural network GAT model.
3. An enterprise risk early warning device, the device comprising:
The receiving module is used for receiving the early warning instruction; wherein, the early warning instruction includes: information of a plurality of entity objects and association relation information among the entity objects;
The drawing module is used for drawing an enterprise association map according to the information of the plurality of entity objects and the association relation information among the entity objects; wherein, the enterprise association graph includes: an edge characterizing a relationship between a node of an entity object and the entity object;
The computing module is used for computing a risk evaluation value of each entity object by utilizing a preset risk evaluation model based on the enterprise association graph;
the early warning module is used for judging whether the risk evaluation value of the entity object is larger than a preset threshold value, and if so, sending a risk early warning message to a designated terminal corresponding to the entity object;
Wherein, the drawing module is used for:
drawing nodes representing the respective entity objects, and setting node attributes for the nodes of each entity object; wherein the node attributes include: a preset potential risk value;
Drawing edges among the nodes according to the association relation information among the entity objects, and setting edge attributes for each edge to form an enterprise association graph; wherein the edge attributes include: the association vector is used for representing the association relation type, the association vector is a vector with 1 XN dimension, and N is the category number of the association relation type;
The computing module is used for:
aiming at a target node, determining a related node connected with the target node through an edge in the enterprise related graph;
Respectively calculating a risk evaluation value of each associated node, and respectively calculating an attention weight value of each associated node to the target node, wherein the attention weight value is specifically as follows: calculating the attention weight W ba of the associated node b to the target node a according to W ba＝softmax(Qa×Rab), wherein Q a is a potential risk value in the node attribute of the target node a, and R ab is an associated vector in the edge attribute of the edge between the target node a and the associated node b;
according to the risk evaluation value of each associated node, the attention weight value of each associated node to the target node and the potential risk value in the node attribute of the target node, the risk evaluation value of the target node is calculated, specifically: according to V a-＝f(Wba×Vb+Wca×Vc + …), calculating a conduction risk value V a- of the associated node to the target node a, wherein f () is a sigmoid function, V b is a risk evaluation value of the associated node b, V c is a risk evaluation value of the associated node c, W ba is an attention weight value of the associated node b to the target node a, W ca is an attention weight value of the associated node c to the target node a, and calculating a risk evaluation value V a+ of the target node a according to V a+＝Va-+Va, wherein V a is a potential risk value in a node attribute according to the target node a.
4. A computer device, the computer device comprising: memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any one of claims 1 to 2 when the computer program is executed.
5. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method according to any one of claims 1 to 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011036277.1A CN112184012B (en) | 2020-09-27 | 2020-09-27 | Enterprise risk early warning method, device, equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011036277.1A CN112184012B (en) | 2020-09-27 | 2020-09-27 | Enterprise risk early warning method, device, equipment and readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112184012A CN112184012A (en) | 2021-01-05 |
| CN112184012B true CN112184012B (en) | 2024-05-31 |
Family
ID=73943691
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011036277.1A Active CN112184012B (en) | 2020-09-27 | 2020-09-27 | Enterprise risk early warning method, device, equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112184012B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112150013A (en) * | 2020-09-27 | 2020-12-29 | 平安资产管理有限责任公司 | Enterprise risk early warning method, device, equipment and readable storage medium |
| CN112965872A (en) * | 2021-02-24 | 2021-06-15 | 刘志欣 | Risk prediction method and device for system cluster nodes, terminal and storage medium |
| CN113094506B (en) * | 2021-04-14 | 2023-08-18 | 每日互动股份有限公司 | Early warning method based on relational graph, computer equipment and storage medium |
| CN113657991A (en) * | 2021-08-12 | 2021-11-16 | 东方微银科技股份有限公司 | Credit risk early warning method, system and storage medium based on graph rule engine |
| CN113642926B (en) * | 2021-08-31 | 2023-08-08 | 上海明略人工智能(集团)有限公司 | Method and device for risk early warning, electronic equipment and storage medium |
| CN116187767B (en) * | 2023-04-26 | 2023-07-14 | 北京网智天元大数据科技有限公司 | Enterprise risk conduction prediction method and system based on graph attention mechanism |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109559192A (en) * | 2018-10-25 | 2019-04-02 | 深圳壹账通智能科技有限公司 | Risk checking method, device, equipment and storage medium based on association map |
| CN109657917A (en) * | 2018-11-19 | 2019-04-19 | 平安科技（深圳）有限公司 | Assess method for prewarning risk, device, computer equipment and the storage medium of object |
| CN109657918A (en) * | 2018-11-19 | 2019-04-19 | 平安科技（深圳）有限公司 | Method for prewarning risk, device and the computer equipment of association assessment object |
| CN109858740A (en) * | 2018-12-21 | 2019-06-07 | 中化资本有限公司 | Appraisal procedure, device, computer equipment and the storage medium of business risk |
| CN110163413A (en) * | 2019-04-15 | 2019-08-23 | 深圳壹账通智能科技有限公司 | Enterprise supervision and method for early warning, device, computer equipment and readable storage medium storing program for executing |
| CN110390465A (en) * | 2019-06-18 | 2019-10-29 | 深圳壹账通智能科技有限公司 | Air control analysis and processing method, device and the computer equipment of business datum |
| CN110738388A (en) * | 2019-09-02 | 2020-01-31 | 深圳壹账通智能科技有限公司 | Method, device, equipment and storage medium for risk conduction of associated map evaluation |
| CN111222681A (en) * | 2019-11-05 | 2020-06-02 | 量子数聚（北京）科技有限公司 | Data processing method, device, equipment and storage medium for enterprise bankruptcy risk prediction |
| CN111402064A (en) * | 2020-06-03 | 2020-07-10 | 天云融创数据科技（北京）有限公司 | Risk value evaluation method and device |
| CN111709661A (en) * | 2020-06-23 | 2020-09-25 | 中国平安财产保险股份有限公司 | Risk processing method, device and equipment for business data and storage medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1361526A1 (en) * | 2002-05-08 | 2003-11-12 | Accenture Global Services GmbH | Electronic data processing system and method of using an electronic processing system for automatically determining a risk indicator value |
| US20040054563A1 (en) * | 2002-09-17 | 2004-03-18 | Douglas William J. | Method for managing enterprise risk |
-
2020
- 2020-09-27 CN CN202011036277.1A patent/CN112184012B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109559192A (en) * | 2018-10-25 | 2019-04-02 | 深圳壹账通智能科技有限公司 | Risk checking method, device, equipment and storage medium based on association map |
| CN109657917A (en) * | 2018-11-19 | 2019-04-19 | 平安科技（深圳）有限公司 | Assess method for prewarning risk, device, computer equipment and the storage medium of object |
| CN109657918A (en) * | 2018-11-19 | 2019-04-19 | 平安科技（深圳）有限公司 | Method for prewarning risk, device and the computer equipment of association assessment object |
| CN109858740A (en) * | 2018-12-21 | 2019-06-07 | 中化资本有限公司 | Appraisal procedure, device, computer equipment and the storage medium of business risk |
| CN110163413A (en) * | 2019-04-15 | 2019-08-23 | 深圳壹账通智能科技有限公司 | Enterprise supervision and method for early warning, device, computer equipment and readable storage medium storing program for executing |
| CN110390465A (en) * | 2019-06-18 | 2019-10-29 | 深圳壹账通智能科技有限公司 | Air control analysis and processing method, device and the computer equipment of business datum |
| CN110738388A (en) * | 2019-09-02 | 2020-01-31 | 深圳壹账通智能科技有限公司 | Method, device, equipment and storage medium for risk conduction of associated map evaluation |
| CN111222681A (en) * | 2019-11-05 | 2020-06-02 | 量子数聚（北京）科技有限公司 | Data processing method, device, equipment and storage medium for enterprise bankruptcy risk prediction |
| CN111402064A (en) * | 2020-06-03 | 2020-07-10 | 天云融创数据科技（北京）有限公司 | Risk value evaluation method and device |
| CN111709661A (en) * | 2020-06-23 | 2020-09-25 | 中国平安财产保险股份有限公司 | Risk processing method, device and equipment for business data and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112184012A (en) | 2021-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112184012B (en) | Enterprise risk early warning method, device, equipment and readable storage medium | |
| US20210264272A1 (en) | Training method and system of neural network model and prediction method and system | |
| WO2020020088A1 (en) | Neural network model training method and system, and prediction method and system | |
| CN112150014B (en) | Enterprise risk early warning method, device, equipment and readable storage medium | |
| CN110751286B (en) | Training method and training system for neural network model | |
| US20230132208A1 (en) | Systems and methods for classifying imbalanced data | |
| US8296205B2 (en) | Connecting decisions through customer transaction profiles | |
| US20200005410A1 (en) | System and Method for Facilitating Legal Review for Commercial Loan Transactions | |
| CN110705719A (en) | Method and apparatus for performing automatic machine learning | |
| US12020257B2 (en) | Generating a fraud prediction utilizing a fraud-prediction machine-learning model | |
| CN112150013A (en) | Enterprise risk early warning method, device, equipment and readable storage medium | |
| CN110751285A (en) | Training method and system and prediction method and system of neural network model | |
| CN113011646A (en) | Data processing method and device and readable storage medium | |
| CN110751287A (en) | Training method and system and prediction method and system of neural network model | |
| CN110858253A (en) | Method and system for executing machine learning under data privacy protection | |
| CN115080858A (en) | Data recommendation method and device under multi-party collaboration scene | |
| Sancetta | Intraday end-of-day volume prediction | |
| CN116703555A (en) | Early warning method, early warning device, electronic equipment and computer readable medium | |
| CN110362981B (en) | Method and system for judging abnormal behavior based on trusted device fingerprint | |
| US20110078071A1 (en) | Prioritizing loans using customer, product and workflow attributes | |
| CN111178535A (en) | Method and device for realizing automatic machine learning | |
| CN113657910B (en) | Real name authentication method, device, electronic equipment and readable storage medium | |
| US20220207409A1 (en) | Timeline reshaping and rescoring | |
| Zhang et al. | Application of BSDE in standard inventory financing loan | |
| CN115374861A (en) | Artificial intelligence based anomaly prediction method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |