CN111598408A - Construction method and application of trade information risk early warning model - Google Patents
Construction method and application of trade information risk early warning model Download PDFInfo
- Publication number
- CN111598408A CN111598408A CN202010326920.8A CN202010326920A CN111598408A CN 111598408 A CN111598408 A CN 111598408A CN 202010326920 A CN202010326920 A CN 202010326920A CN 111598408 A CN111598408 A CN 111598408A
- Authority
- CN
- China
- Prior art keywords
- entity
- nodes
- risk
- early warning
- edges
- 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.)
- Granted
Links
Images
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
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention discloses a construction method and application of a trade information risk early warning model, wherein the construction method comprises the following steps: s1, constructing a complex network: extracting effective entity information from a historical commodity information table, taking each effective entity information as an entity node, connecting two associated entity nodes to form an edge, wherein the entity nodes and the edge form a complex network structure; s2: quantizing the complex network: counting spot check data of a historical commodity information table to calculate a problem rate, and quantifying risk values of entity nodes and edges by taking the problem rate as a quantitative characteristic of the relation between the entity nodes and the edges; s3: constructing a risk early warning model: and weighting the risk values of the entity nodes and the edges according to the spot check data of the historical commodity information table to obtain a weight function, and then weighting and summing the risk values of the entity nodes and the edges to obtain a risk early warning model. The invention solves the problems of low sampling inspection accuracy and large sampling inspection workload in the prior art.
Description
Technical Field
The invention relates to the technical field of sampling inspection of complex trade commodities, in particular to a construction method and application of a trade information risk early warning model.
Background
During the transportation of the traded goods, the transportation traffic (trucks, trains, cargo ships, airplanes) reports an information table, and the table records the relevant information of the transported goods, such as trade names, weight of the goods, manufacturing enterprises, cargo owner units, logistics units, source places, destinations, etc. If the transported goods do not accord with the reported information (for example, the actual weight or number of the goods do not accord with the reported information, other goods which are not reported are carried along, illegal goods are transported, and the like), the abnormity needs to be dealt with in time. Therefore, when arriving at the destination, the inspector can determine whether the various transported goods are abnormal according to the reported information, the actual inspection result, the local policy and the like. Along with trade transportation in various places is more and more frequent and is limited by time cost, inspectors cannot inspect all commodities one by one and only can carry out selective inspection, although the inspectors with rich experience can carry out selective inspection more effectively according to experience, the time and the labor are still wasted, the experience of the selective inspection work is difficult to be solidified, and the problem that how to improve the inspection efficiency (to inspect abnormal commodities and abnormal enterprises as fast as possible) and reduce the labor cost is to be solved urgently is solved.
The accuracy of modeling by using machine learning and deep learning methods is influenced by various factors, for example, when data are not greatly correlated in each dimension and positive and negative samples of the data are extremely unbalanced, the methods may not learn effective characteristics, and the economy and the real-time performance of the model are poor, so that a large amount of calculation power is usually consumed for training. In the trade transportation of commodities, the data dimension information contained in the reported information table is too discrete and needs manual integration. For example, the information table includes the business establishment time, and if it is desired to quantify the business establishment time, the business establishment time needs to be subtracted from the current time. And the data of different information tables are often related, the trade activity is influenced by multiple parties, for example, the interest relations among production enterprises, transport companies, agents and the like are relatively complex, so that a single information table is directly used as input to train a machine learning or deep learning model, the accuracy of predicting problem commodities or problem enterprises is not higher than that of manual spot inspection, and the effect is not ideal.
Disclosure of Invention
The invention aims to provide a construction method of a trade information risk early warning model, which is used for performing spot check on trade commodities and solves the problems of low spot check accuracy and large spot check workload in the prior art.
The invention is realized by the following technical scheme:
a construction method of a trade information risk early warning model comprises the following steps:
s1, constructing a complex network: extracting effective entity information from a historical commodity information table, taking each effective entity information as an entity node, and forming an edge by connecting two associated entity nodes, wherein the connecting line between the entity nodes in the same commodity information table is an internal edge of the table, the connecting line between the entity nodes in different commodity information tables is an external edge of the table, and the entity nodes and the edges form a complex network structure;
s2: quantizing the complex network: counting spot check data of a historical commodity information table to calculate a problem rate, and quantifying risk values of entity nodes and edges by taking the problem rate as a quantitative characteristic of the relation between the entity nodes and the edges;
s3: constructing a risk early warning model: weighting the risk values of the entity nodes and the edges according to the spot check data of the historical commodity information table to obtain a weight function, then weighting and summing the risk values of the entity nodes and the edges to obtain a risk early warning model, wherein the weight is an importance coefficient of a risk item, and the weight function is adjusted according to the spot check data of the historical commodity information table.
The present invention constructs nodes and edges of a complex network based on historical inspection data. The entity nodes of the complex network represent certain entities, such as commodity types, arrival time, enterprise information, contact calls, inspection results and the like; the edge of the complex network represents the relationship between the entity nodes at two ends of the complex network; the commodity information table is used as a basic unit for explanation. When a complex network structure is built, each commodity information table extracts information with the same dimensionality as an entity node, the network structures of the information tables are connected (at least connected with table IDs), and the mutual association specifically means that whether a certain relation exists between 2 entity nodes is judged according to actual data, for example, a commodity E is produced by an O company, and then the commodity E is associated with the O company.
The invention can carry out risk assessment, namely, the inspection result and the prediction result are counted and displayed according to any continuous time interval, including problem items (such as weight and reporting inconsistency) of the actual inspection result, the prediction result of the commodity information table, and risk indexes of certain entity nodes or edges (such as risk distribution of enterprise nodes or commodity nodes, risk distribution of enterprise and enterprise edge relations or commodity and commodity edge relations), so that staff can analyze the change trend of abnormal information to guide subsequent inspection work. In actual use, data which is too long (for example, 5 years ago) is abandoned, the latest data is added into the model, and the periodic update of the model is realized, wherein the method comprises the steps of adding and deleting nodes and edges in a complex network, updating risk values, updating weights and the like.
Experiments show that the prediction accuracy of the model is higher than that of manual screening, which shows that the modeling method based on the complex network has certain risk prediction capability, and the change trend of abnormal information can be found through risk index analysis.
Further, the process of extracting valid entity information in step S1 is as follows:
selecting N commodity information tables from the historical data, and selecting the commodity information table S for each commoditynIf the K-dimensional effective entity information is taken as an entity node of the complex network, N · K entity nodes are constructed in the complex network.
The invention constructs the edge relation in the same way for each commodity information table (as shown in figure 1):
the K nodes in the commodity information table should be communicated and can be connected randomly in principle, but should have a certain actual meaning, for example, the commodity nodes are connected with nodes of a production enterprise to indicate that the commodity and the enterprise have relationships of production, purchase, transportation and the like. E to be constructed for a single information table1The continuous edge with practical significance is called the edge inside the table, and K-1 is more than or equal to E in each commodity information table1Less than or equal to K (K-1)2 edges; namely N.E is constructed in a complex network1The strip inner edge.
Further, in step S1, the external edge in the table is a connection line between entity nodes of the same type in different commodity information tables or a connection line between entity nodes of different types in different commodity information tables.
For the same kind of information tables of different commoditiesType nodes are bordered as shown by the type A and type B nodes in FIG. 2 (i.e., A1And A2Are connected to each other, B1And B2Connecting; different types of nodes can be connected, whether the connection is related to actual data or not), and the edge represents the relationship of the same type of entity nodes at the two ends. For example, the edges of the enterprise nodes of the different merchandise information tables indicate that a certain relationship exists between the enterprise and the enterprise, such as a long-term cooperation relationship or a simultaneous transportation relationship. Thus, in a complex network, E is constructed2And edges among different information tables of the same type of nodes are arranged.
Connecting edges are performed on different types of nodes of different commodity information tables, as shown by the nodes of class C and class D in FIG. 2 (i.e. class C1And D2Are connected to each other, C2And D1Connected, connected or not related to actual data), an edge represents that different types of entity nodes at two ends of the edge have certain relation. For example, the same natural person acts as a legal representative of a plurality of enterprises, an enterprise produces different commodities, and the like. Thus, constructing E3An edge.
Further, the entity nodes in step S1 include a commodity node, a production enterprise node, a carrier node, an agent node, and a legal representative node.
Further, the problem rates in step S2 include an entity node problem rate, a table inside edge problem rate, and a table outside edge problem rate.
Further, the problem rate of the entity node is the ratio of the number of abnormal inspection results of the entity node to the total number of inspection results.
For example, for the commodity node itself, through the statistical historical inspection data, each commodity has a problem rate, and the problem rate is defined as
R (G) is the risk value of commodity G, and similarly, production enterprises, transportation companies, agent companies, legal representatives and the like can be used as the problem rate of the nodes in the complex network through the problem rate in the information table.
Further, the problem rate of the edges inside the table is calculated by using the entity nodes at the two ends, and when the entity nodes at the two ends are commodities and production enterprises, the calculation model of the problem rate of the edges inside the table is as follows:
further, the problem rate of the table outer edge is calculated by using the entity nodes at the two ends, and when the entity nodes at the two ends represent two commodities, the calculation model of the problem rate of the table outer edge is as follows:
further, the risk early warning model is as follows:
wherein S is a commodity information table, K is the number of entity nodes in the commodity information table S, E1Number of edges in the table, EoutFor the number of table outer edges, W is the weighting function and R is the risk value.
The application of the trade information risk early warning model is characterized in that the constructed risk early warning model is used for commodity information risk prediction in new transportation behaviors:
firstly, extracting entity nodes and edges in a commodity information table according to a modeling method for constructing a risk early warning model; matching in the complex network model, and assigning risk values of the same entity node and the same edge in the complex network model to a commodity information table to be predicted; and predicting the commodity information risk according to a weighted summation mode in the risk early warning model.
According to the invention, the risk value of the commodity information table is predicted by utilizing the constructed risk early warning model, so that the actual sampling inspection amount can be reduced, the hit rate of abnormal commodities is improved, the working efficiency of inspectors is improved, the characteristics are solidified to a certain extent by the risk index items, and the personnel cultivation cost is reduced.
For example:
let the commodity information table to be predicted be SnewA certain node thereof is VaAnd matching in the complex network model. If there is also a node V in the modelbAnd node VaExactly the same (i.e., node type is the same as content, e.g., same business), then the business risk value R (V)a)＝R(Vb) If V isaIf not matched, then R (V)a)＝Cv，CvIs a constant.
Information table S for setting commodity to be predictednewA certain side of is EaSimilarly, if there is an edge E in the complex networkbAnd EaIdentical (i.e. node type and content at both ends of the edge are denoted by R (S)new) And the relationship represented by the edge is the same), then R (E)a)＝R(Eb) If E isaNot matched, then R (E)a)＝Ce，CeIs a constant.
According to the weighting summation mode of the complex network model, the weight function W is the same as the model, and a new information table S is obtainednewRisk prediction of (a):
then, a threshold T is set, when R (S)new) When the risk value is larger than T, the model reminds inspectors that the transported goods have higher risk, and meanwhile, the risk values of the entity nodes and the edge relations of the goods information table are fed back to assist manual screening.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. the invention adopts a complex network method to carry out data modeling on trade risks, utilizes effective entity information of a commodity information table to construct entity nodes of the complex network, utilizes the relation between the entity nodes to construct edges of the complex network, and utilizes historical data to quantify the risk values of the nodes and the edges, namely effective characteristics are designed, the characteristics of the data entities and the relation between the data can be deeply expressed, and the risk characteristics of the commodity information table can be better focused. The risk value of the commodity information table is predicted by utilizing the constructed risk early warning model, the actual sampling inspection amount can be reduced, the hit rate of abnormal commodities is improved, the working efficiency of inspectors is improved, and the problems that the sampling inspection accuracy rate is low and the sampling inspection workload is large in the prior art are solved.
2. According to the invention, the characteristics are solidified to a certain extent through the risk index items in the constructed risk early warning model, and the personnel cultivation cost is reduced.
3. The invention can count and display all the inspection results, the inspection abnormal results and the model prediction results according to any continuous time, and count the abnormal items of the inspection, the overall risk value of the model prediction and the risk indexes (such as the risk indexes of enterprises and the risk indexes of commodities) of various nodes and sides, thereby forming a visual risk evaluation table for the staff to analyze the abnormal trend of trade transportation; in addition, the database of the system stores a new commodity information table, the model needs to update parameters regularly, and the method mainly comprises the following steps: 1) adding and deleting entity nodes and edges; 2) re-counting the problem rates of the entity nodes and edges, and updating the risk values; 3) and updating the weight function.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
FIG. 1 is a schematic diagram of a network structure of a merchandise information table;
FIG. 2 is a schematic diagram showing a node edge connecting mode of two commodity information tables;
fig. 3 is a schematic view of a risk early warning process of trade information.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.
Example 1:
as shown in fig. 3, a method for constructing a trade information risk early warning model includes the following steps:
s1, constructing a complex network: extracting effective entity information from a historical commodity information table, taking each effective entity information as an entity node, and forming an edge by connecting two associated entity nodes, wherein the connecting line between the entity nodes in the same commodity information table is an internal edge of the table, the connecting line between the entity nodes in different commodity information tables is an external edge of the table, and the entity nodes and the edges form a complex network;
specifically, the method comprises the following steps:
s11, constructing entity nodes of the complex network:
selecting N commodity information tables from the historical data, and selecting N commodity information tables S for each commodity information tablenExtracting effective entity information of K dimensions, and constructing N.K entity nodes in the complex network by taking the effective entity information as entity nodes of the complex network;
s12 constructs edges of the complex network:
s121, constructing the edge relation of each item information table in the same manner, as shown in fig. 1.
The K nodes in the commodity information table should be connected and can be connected arbitrarily in principle, but the connected edges should have certain practical meanings, for example, the commodity nodes are connected with the nodes of a production enterprise to indicate that the commodity and the enterprise have relationships of production, purchase or transportation and the like. Let E construct a single information table1The continuous edge with practical significance is called as the edge inside the table, and K-1 is more than or equal to E in each commodity information table1Less than or equal to K (K-1)/2 sides. Thus, N.E is constructed in the complex network1An inner edge of the strip;
s122, connecting edges to the same type nodes of different commodity information tables, as shown in the a type and B type nodes in fig. 2, where the edges represent the relationship between the same type entity nodes at both ends. For example, the edges of the enterprise nodes of the different merchandise information tables indicate that a certain relationship exists between the enterprise and the enterprise, such as a long-term cooperation relationship or a simultaneous transportation relationship. Thus, in a complex network, E is constructed2Edges among different information tables of the same type of nodes are listed;
s123, for different commodity information tableThe different types of nodes in (2) are connected, as shown in the nodes of class C and class D in fig. 2, an edge represents that there is some relationship between different types of entity nodes at both ends. For example, the same natural person acts as a legal representative of a plurality of enterprises, an enterprise produces different commodities, and the like. Thus, constructing E3A side;
the structure of the complex network has N.K nodes, N.E1+E2+E3Edge, E1Is an internal edge of the watch, E2And E3The outer edge of the table.
S2: quantizing the complex network: counting spot check data of a historical commodity information table to calculate a problem rate, and quantifying risk values of entity nodes and edges by taking the problem rate as a quantitative characteristic of the relation between the entity nodes and the edges;
the main purpose of this step is to quantify the structural features of the complex network, i.e. to assign scores (or risk values) to the nodes and edges of the complex network. The risk value of the node represents the risk coefficient of the entity node, and is related to all nodes of the type in the network, and if the problem rate in the historical inspection data of a certain commodity is high, the risk value of the commodity node of the type is relatively large; the risk value of the edge is related to the node types of the nodes at the two ends and the relationship between the nodes, and if a certain commodity produced by a certain enterprise has high problem rate and other produced commodities hardly have problems, a relatively large risk value is given to the edge connecting the enterprise node and the commodity node.
In the manual inspection of commodity transportation, the commodity or the enterprise with problems in the historical inspection is selected for inspection in a targeted manner mainly according to the historical inspection result, and the problems are more likely to occur in the future trade of the commodity or the enterprise with problems in the historical inspection. Therefore, in the complex network characteristic modeling, the risk values of the nodes and the edges are mainly quantified by using the historical inspection results.
For example, for the commodity node itself, through the statistical historical inspection data, each commodity has a problem rate, and the problem rate is defined as
R (G) is the risk value of the commodity G, and similarly, a production enterprise, a transportation company, an agent company, a legal representative and the like can be used as the problem rate of the node in the complex network through the problem rate in the information table.
For the table internal edge of the complex network, the data is checked through statistical history, and the problem rate is also used as the risk value of the edge. For example, if the edge structure represents the enterprise producing the commodity, then the definition is defined
By using
The problem rate when the product G and the enterprise M appear in one product information table at the same time is shown, and the value is given as a risk value on the corresponding side.
For the table external edge of the complex network, nodes at two ends of the edge are arranged to represent certain two commodities, and the edge structure represents that the two commodities are transported simultaneously, then
In fact, the information table is submitted in the unit of transportation, and a plurality of commodities are possible in one transportation, and sometimes, certain commodities are always transported together, so that the commodities have certain relation. When the external edges of the table are connected with different types of nodes, the calculation principle of the risk values is the same.
S3: constructing a risk early warning model: weighting the risk values of the body nodes and the edges according to the spot check data of the historical commodity information table to obtain a weight function, then weighting and summing the risk values of the entity nodes and the edges to obtain a risk early warning model, wherein the weight is an importance coefficient of a risk item, and the weight function is adjusted according to the spot check data of the historical commodity information table.
For a commodity information table S, it has K nodes and E1The internal side of the bar is provided with EoutExterior edge of bar watch and watchThe internal nodes are connected, and the nodes and the edges have respective risk values. The above are weighted according to type to obtain a weight function W (for example, if the weight of the a-type node is a, W (a) ═ a, the weight can be understood as the importance coefficient of the risk term), and then the risk values of the node and the edge are weighted and summed to obtain a value r (S) as the risk value of the product information table S:
wherein S is a commodity information table, K is the number of entity nodes in the commodity information table S, E1Number of edges in the table, EoutFor the number of table outer edges, W is the weighting function and R is the risk value.
The weighting function W is adjusted in accordance with the historical data, in principle, so that the risk value of the problematic item information table is as high as possible and the risk value of the normal item information table is as low as possible.
The application of the trade information risk early warning model is to apply the risk early warning model constructed in the embodiment 1 to commodity information risk prediction in new transportation behaviors:
a new information table S is provided, as shown in fig. 1, wherein the node meaning, the risk value obtained by matching, and the weight are:
a-commercial value, 0.10, 0.10; b-transportation agency, 0.20, 0.15;
c-manufacturing company, 0.16, 0.05; d-manufacturing site, 0.22, 0.10;
e-commercial weight, 0.05, 0.05; variety F, 0.08, 0.05;
g-apple, 0.15, 0.10; h-business registered capital, 0.20, 0.15;
I-Enterprise legal representatives, 0.30, 0.25;
the risk value and weight of the edge (the detailed meaning is omitted) are:
(A,E)，0.20，0.10；(E,G)，0.03，0.05；
(F,G)，0.04，0.30；(B,G)，0.25，0.20；
(B,I)，0.16，0.15；(C,G)，0.04，0.05；
(C,D)，0.20，0.10；(C,H)，0.35，0.05；
accordingly, the risk value R of the information table is calculated to be 0.3435 according to the model; the threshold value T is set to 0.3, and R > T, so that the model predicts that the information table is high risk, and feeds back a risk index (i.e., a score of each node and edge) to an inspector, and the inspector determines whether to carefully inspect the commodity according to actual conditions and a model prediction result.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. A construction method of a trade information risk early warning model is characterized by comprising the following steps:
s1, constructing a complex network: extracting effective entity information from a historical commodity information table, taking each effective entity information as an entity node, and forming an edge by connecting two associated entity nodes, wherein the connecting line between the entity nodes in the same commodity information table is an internal edge of the table, the connecting line between the entity nodes in different commodity information tables is an external edge of the table, and the entity nodes and the edges form a complex network structure;
s2: quantizing the complex network: counting spot check data of a historical commodity information table to calculate a problem rate, and quantifying risk values of entity nodes and edges by taking the problem rate as a quantitative characteristic of the relation between the entity nodes and the edges;
s3: constructing a risk early warning model: weighting the risk values of the entity nodes and the edges according to the spot check data of the historical commodity information table to obtain a weight function, then weighting and summing the risk values of the entity nodes and the edges to obtain a risk early warning model, wherein the weight is an importance coefficient of a risk item, and the weight function is adjusted according to the spot check data of the historical commodity information table.
2. The method for constructing the trade information risk early warning model according to claim 1, wherein the process of extracting the valid entity information in the step S1 is as follows:
selecting N commodity information tables from the historical data, and selecting the commodity information table S for each commoditynIf the K-dimensional effective entity information is taken as an entity node of the complex network, N · K entity nodes are constructed in the complex network.
3. The method for constructing the trade information risk early warning model according to claim 1, wherein the external edges in the table in step S1 are the same type of connection between the entity nodes in different commodity information tables or different types of connection between the entity nodes in different commodity information tables.
4. The method for constructing a risk early warning model of trade information as claimed in claim 1, wherein said entity nodes in step S1 include commodity nodes, production enterprise nodes, transportation company nodes, agent company nodes and legal representative nodes.
5. The method for constructing the trade information risk early warning model according to claim 1, wherein the problem rates in the step S2 include a solid node problem rate, a table inside edge problem rate and a table outside edge problem rate.
6. The method as claimed in claim 5, wherein the problem rate of the entity node is a ratio of the number of abnormal inspection results to the total number of inspection results of the entity node.
7. The method for constructing the trade information risk early warning model according to claim 5, wherein the problem rate of the edge inside the table is calculated by using the entity nodes at two ends, and when the entity nodes at two ends are commodities and production enterprises, the calculation model of the problem rate of the edge inside the table is as follows:
8. the method as claimed in claim 5, wherein the problem rate of the table outer edge is calculated by using the entity nodes at the two ends, and when the entity nodes at the two ends represent two commodities, the calculation model of the problem rate of the table outer edge is as follows:
9. the method for constructing the trade information risk early warning model according to any one of claims 1 to 8, wherein the risk early warning model is as follows:
wherein S is a commodity information table, K is the number of entity nodes in the commodity information table S, E1Number of edges in the table, EoutFor the number of table outer edges, W is the weighting function and R is the risk value.
10. Use of a trade information risk early warning model constructed according to any one of claims 1 to 9 for commodity information risk prediction in new transportation behavior:
firstly, extracting entity nodes and edges in a commodity information table according to a modeling method for constructing a risk early warning model; matching in the complex network model, and assigning risk values of the same entity node and the same edge in the complex network model to a commodity information table to be predicted; and predicting the commodity information risk according to a weighted summation mode in the risk early warning model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010326920.8A CN111598408B (en) | 2020-04-23 | 2020-04-23 | Construction method and application of trade information risk early warning model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010326920.8A CN111598408B (en) | 2020-04-23 | 2020-04-23 | Construction method and application of trade information risk early warning model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111598408A true CN111598408A (en) | 2020-08-28 |
| CN111598408B CN111598408B (en) | 2023-04-18 |
Family
ID=72192402
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010326920.8A Active CN111598408B (en) | 2020-04-23 | 2020-04-23 | Construction method and application of trade information risk early warning model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111598408B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113610447A (en) * | 2021-10-09 | 2021-11-05 | 北京奇虎科技有限公司 | Method, device, storage medium and apparatus for detecting trade compliance |
Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006244196A (en) * | 2005-03-04 | 2006-09-14 | Osamu Uno | Support system for managing international transportation risk |
| US20140195984A1 (en) * | 2013-01-07 | 2014-07-10 | Northeastern University | Analytic frameworks for persons of interest |
| US9100430B1 (en) * | 2014-12-29 | 2015-08-04 | Palantir Technologies Inc. | Systems for network risk assessment including processing of user access rights associated with a network of devices |
| US20150278729A1 (en) * | 2014-03-28 | 2015-10-01 | International Business Machines Corporation | Cognitive scoring of asset risk based on predictive propagation of security-related events |
| US20160004830A1 (en) * | 2012-05-25 | 2016-01-07 | Echometrics Cardiologists, Pc | Determining disease state of a patient by mapping a topological module representing the disease, and using a weighted average of node data |
| CN107194623A (en) * | 2017-07-20 | 2017-09-22 | 深圳市分期乐网络科技有限公司 | A kind of discovery method and device of clique's fraud |
| CN108830507A (en) * | 2018-06-29 | 2018-11-16 | 成都数之联科技有限公司 | A kind of food safety risk method for early warning |
| CN108898300A (en) * | 2018-06-21 | 2018-11-27 | 南京邮电大学 | The construction method of supply chain network risk cascade model |
| CN109598552A (en) * | 2018-11-28 | 2019-04-09 | 东南大学 | A kind of cross-border electricity trading market characteristics analysis method based on Complex Networks Theory |
| CN109639670A (en) * | 2018-12-10 | 2019-04-16 | 北京威努特技术有限公司 | A kind of industry control network security postures quantitative estimation method of knowledge based map |
| WO2019114434A1 (en) * | 2017-12-15 | 2019-06-20 | 阿里巴巴集团控股有限公司 | Graphical structure model-based method for transaction risk control, and device and equipment |
| CN109934697A (en) * | 2017-12-15 | 2019-06-25 | 阿里巴巴集团控股有限公司 | A kind of credit risk control method, device and equipment based on graph structure model |
| CN110321460A (en) * | 2019-07-01 | 2019-10-11 | 成都数之联科技有限公司 | A kind of food safety association map construction method and system |
| CN110766321A (en) * | 2019-10-23 | 2020-02-07 | 浙江工商大学 | Emergency logistics network risk propagation method and system based on material reliability |
| CN110852584A (en) * | 2019-10-30 | 2020-02-28 | 北京航空航天大学 | Risk identification method based on complex network analysis |
| CN110956223A (en) * | 2018-12-24 | 2020-04-03 | 宁波大学 | Method for summarizing complex network diagram and application of complex network diagram in protein relation network |
| CN110991913A (en) * | 2019-12-09 | 2020-04-10 | 南京航空航天大学 | Busy airport peak time congestion risk analysis method |
-
2020
- 2020-04-23 CN CN202010326920.8A patent/CN111598408B/en active Active
Patent Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006244196A (en) * | 2005-03-04 | 2006-09-14 | Osamu Uno | Support system for managing international transportation risk |
| US20160004830A1 (en) * | 2012-05-25 | 2016-01-07 | Echometrics Cardiologists, Pc | Determining disease state of a patient by mapping a topological module representing the disease, and using a weighted average of node data |
| US20140195984A1 (en) * | 2013-01-07 | 2014-07-10 | Northeastern University | Analytic frameworks for persons of interest |
| US20150278729A1 (en) * | 2014-03-28 | 2015-10-01 | International Business Machines Corporation | Cognitive scoring of asset risk based on predictive propagation of security-related events |
| US9100430B1 (en) * | 2014-12-29 | 2015-08-04 | Palantir Technologies Inc. | Systems for network risk assessment including processing of user access rights associated with a network of devices |
| CN107194623A (en) * | 2017-07-20 | 2017-09-22 | 深圳市分期乐网络科技有限公司 | A kind of discovery method and device of clique's fraud |
| WO2019114434A1 (en) * | 2017-12-15 | 2019-06-20 | 阿里巴巴集团控股有限公司 | Graphical structure model-based method for transaction risk control, and device and equipment |
| CN109934697A (en) * | 2017-12-15 | 2019-06-25 | 阿里巴巴集团控股有限公司 | A kind of credit risk control method, device and equipment based on graph structure model |
| CN108898300A (en) * | 2018-06-21 | 2018-11-27 | 南京邮电大学 | The construction method of supply chain network risk cascade model |
| CN108830507A (en) * | 2018-06-29 | 2018-11-16 | 成都数之联科技有限公司 | A kind of food safety risk method for early warning |
| CN109598552A (en) * | 2018-11-28 | 2019-04-09 | 东南大学 | A kind of cross-border electricity trading market characteristics analysis method based on Complex Networks Theory |
| CN109639670A (en) * | 2018-12-10 | 2019-04-16 | 北京威努特技术有限公司 | A kind of industry control network security postures quantitative estimation method of knowledge based map |
| CN110956223A (en) * | 2018-12-24 | 2020-04-03 | 宁波大学 | Method for summarizing complex network diagram and application of complex network diagram in protein relation network |
| CN110321460A (en) * | 2019-07-01 | 2019-10-11 | 成都数之联科技有限公司 | A kind of food safety association map construction method and system |
| CN110766321A (en) * | 2019-10-23 | 2020-02-07 | 浙江工商大学 | Emergency logistics network risk propagation method and system based on material reliability |
| CN110852584A (en) * | 2019-10-30 | 2020-02-28 | 北京航空航天大学 | Risk identification method based on complex network analysis |
| CN110991913A (en) * | 2019-12-09 | 2020-04-10 | 南京航空航天大学 | Busy airport peak time congestion risk analysis method |
Non-Patent Citations (1)
| Title |
|---|
| 姜海洋;周宽久;林林;郝佳亮;郭铖;: "战时运输风险评估模型及其路径优化研究" * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113610447A (en) * | 2021-10-09 | 2021-11-05 | 北京奇虎科技有限公司 | Method, device, storage medium and apparatus for detecting trade compliance |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111598408B (en) | 2023-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Samoilenko et al. | Using Data Envelopment Analysis (DEA) for monitoring efficiency-based performance of productivity-driven organizations: Design and implementation of a decision support system | |
| Cullinane et al. | The efficiency analysis of container port production using DEA panel data approaches | |
| Shevchenko et al. | Multi‐attribute analysis of investments risk alternatives in construction | |
| Wibowo et al. | Supply chain management strategy for recycled materials to support sustainable construction | |
| Bititci et al. | Driving continuous improvement | |
| Dinlersoz et al. | The Rise and Fall of Unions in the US | |
| Khalili et al. | Performance evaluation in green supply chain using BSC, DEA and data mining | |
| CN113869768A (en) | Method, device, equipment and readable medium for evaluating strength of industrial chain | |
| CN115018384A (en) | Building site security risk assessment method and system | |
| CN111598408B (en) | Construction method and application of trade information risk early warning model | |
| CN117217634B (en) | Enterprise cooperation community discovery method based on complex network | |
| Zaitsev et al. | A methodology to assess innovation performance in digital economy. | |
| HOANG | Fuzzy-AHP application in analyzing the factors affecting quality of rural labor | |
| Noveiri et al. | Multi-period sustainability performance in the presence of discrete and bounded measures | |
| Gaebert et al. | Efficiency development in the German pharmaceutical market | |
| Link | Methods for evaluating the return on R&D investments | |
| Küfner et al. | Analysing establishment survey non-response using administrative data and machine learning | |
| CN114595945A (en) | Manufacturing process-oriented health degree evaluation method for MBOM (moving object model) | |
| Mahalik et al. | Efficiency Measurement Using DEA and AHP: a case study on Indian ports. | |
| Lu | Spatial economics model predicting transport volume | |
| CN111860936A (en) | Method for predicting defects of office business process | |
| Tong et al. | The evaluation of lean logistics performance based on balanced score card and unascertained sets | |
| Boonsothonsatit | Order Penetration point-based cAusal Linkages (OPAL) to systematically evaluate supply chain performance | |
| Liu et al. | Forecast of Line-level Logistics Cargo Volume Based on Hybrid Dual Neural Network Model | |
| Paklina et al. | Intangible-intensive strategies of Russian companies |
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 | ||
| CB02 | Change of applicant information |
Address after: 610000 No. 270, floor 2, No. 8, Jinxiu street, Wuhou District, Chengdu, Sichuan Applicant after: Chengdu shuzhilian Technology Co.,Ltd. Address before: 610000 No.2, 4th floor, building 1, Jule Road intersection, West 1st section of 1st ring road, Wuhou District, Chengdu City, Sichuan Province Applicant before: CHENGDU SHUZHILIAN TECHNOLOGY Co.,Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |