CN114358607A - Risk monitoring method and device - Google Patents
Risk monitoring method and device Download PDFInfo
- Publication number
- CN114358607A CN114358607A CN202210012741.6A CN202210012741A CN114358607A CN 114358607 A CN114358607 A CN 114358607A CN 202210012741 A CN202210012741 A CN 202210012741A CN 114358607 A CN114358607 A CN 114358607A
- Authority
- CN
- China
- Prior art keywords
- data
- risk monitoring
- knowledge graph
- dimension
- enterprise
- 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.)
- Pending
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 209
- 238000000034 method Methods 0.000 title claims abstract description 64
- 230000004927 fusion Effects 0.000 claims abstract description 50
- 238000012806 monitoring device Methods 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims description 19
- 238000003062 neural network model Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 15
- 230000003993 interaction Effects 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 230000007547 defect Effects 0.000 abstract description 5
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 12
- 230000006870 function Effects 0.000 description 7
- 239000013598 vector Substances 0.000 description 7
- 238000013528 artificial neural network Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 238000010276 construction Methods 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000002457 bidirectional effect Effects 0.000 description 3
- 230000006855 networking Effects 0.000 description 3
- 230000000306 recurrent effect Effects 0.000 description 3
- 238000011835 investigation Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000012550 audit Methods 0.000 description 1
- 230000008033 biological extinction Effects 0.000 description 1
- 230000003925 brain function Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000009193 crawling Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000013077 scoring method Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a risk monitoring method and device, relates to the technical field of artificial intelligence, and can be used in the financial field or other technical fields. The method comprises the following steps: and inputting the fusion data to a preset risk monitoring model, and taking an output result of the preset risk monitoring model as a risk monitoring result. The device performs the above method. The risk monitoring method and the risk monitoring device provided by the embodiment of the invention not only can reduce the labor cost and improve the efficiency, but also can overcome the defect of limited data dimension, and can comprehensively perform risk early warning aiming at the service characteristics.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a risk monitoring method and device.
Background
With the development of data technology, massive data is generated, and meanwhile, data risks are also accompanied. The technology of present automatic monitoring risk indicator is more and more extensive, has had risk early warning correlation technique, nevertheless has not had the shortcoming including: the concerned data has limited dimensionality and can not comprehensively carry out risk early warning aiming at the service characteristics; in addition, risk early warning needs to be with the help of the manpower, leads to too high, the inefficiency of human cost.
Disclosure of Invention
To solve the problems in the prior art, embodiments of the present invention provide a risk monitoring method and apparatus, which can at least partially solve the problems in the prior art.
In one aspect, the present invention provides a risk monitoring method, including:
inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data;
and taking the output result of the preset risk monitoring model as a risk monitoring result.
Wherein the neural network model is obtained by combining BRNN and LSTM.
Wherein the risk monitoring results correspond to customer dimensions, product dimensions, and enterprise dimensions, respectively; correspondingly, after the step of taking the output result of the preset risk monitoring model as the risk monitoring result, the risk monitoring method further includes:
and if at least one of the client dimension risk monitoring result, the product dimension risk monitoring result and the enterprise dimension risk monitoring result has a risk, generating a risk early warning message corresponding to the at least one risk monitoring result having the risk.
Wherein obtaining the fused data comprises:
respectively constructing customer information map data, product information map data and enterprise information map data;
respectively constructing customer risk monitoring map data, product risk monitoring map data and enterprise risk monitoring map data according to risk monitoring threshold values of data items respectively corresponding to the customer information map data, the product information map data and the enterprise information map data and corresponding data items;
and fusing the customer information map data, the customer risk monitoring map data, the product information map data, the product risk monitoring map data, the enterprise information map data and the enterprise risk monitoring map data respectively to obtain the customer dimension comprehensive knowledge map data, the product dimension comprehensive knowledge map data and the enterprise dimension comprehensive knowledge map data.
Wherein, after the step of obtaining the fused data, the risk monitoring method further comprises:
and calculating the node weights in the customer dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data by using a utilization centrality calculation method.
Wherein, after the step of calculating the node weights in the customer dimension integrated knowledge-graph data, the product dimension integrated knowledge-graph data and the enterprise dimension integrated knowledge-graph data by the utilization centrality calculation method, the risk monitoring method further comprises:
endowing the node weight to entities in the dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data, and performing vectorization expression on the dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data which are endowed with the node weight;
and performing operation processing on the client dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data which are represented in the quantitative manner by using a MainfoldE algorithm.
Wherein the risk monitoring method further comprises:
inputting the fusion data which is subjected to the operation processing and is expressed in a vectorization mode to the preset risk monitoring model, and continuing to execute the subsequent steps.
Wherein, obtaining the enterprise risk monitoring atlas data comprises:
acquiring enterprise information, identifying the text content of the enterprise information, and respectively matching the text content with words in a preset positive public opinion library and a preset negative public opinion library;
determining the type of the text content according to the matching result; the type of the text content comprises positive content or negative content;
traversing all the text contents, and acquiring the number of the text contents respectively corresponding to the positive content or the negative content according to the type of the text contents of each text content;
if the number of the text contents corresponding to the positive contents is less than the number of the text contents corresponding to the negative contents, determining that the enterprise information is negative enterprise information;
and acquiring enterprise interaction information corresponding to the negative enterprise information, and if the statistical value of at least one item of interaction index data in the enterprise interaction information is larger than a preset statistical data threshold value, taking the negative enterprise information as the enterprise risk monitoring map data.
In one aspect, the present invention provides a risk monitoring device, including:
the input unit is used for inputting the fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data;
and the monitoring unit is used for taking the output result of the preset risk monitoring model as a risk monitoring result.
In another aspect, an embodiment of the present invention provides an electronic device, including: a processor, a memory, and a bus, wherein,
the processor and the memory are communicated with each other through the bus;
the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform a method comprising:
inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data;
and taking the output result of the preset risk monitoring model as a risk monitoring result.
An embodiment of the present invention provides a non-transitory computer-readable storage medium, including:
the non-transitory computer readable storage medium stores computer instructions that cause the computer to perform a method comprising:
inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data;
and taking the output result of the preset risk monitoring model as a risk monitoring result.
According to the risk monitoring method and device provided by the embodiment of the invention, the fusion data is input to the preset risk monitoring model, and the output result of the preset risk monitoring model is used as the risk monitoring result, so that the labor cost can be reduced, the efficiency can be improved, the defect of limited data dimension can be overcome, and the risk early warning is comprehensively carried out according to the service characteristics.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts. In the drawings:
fig. 1 is a schematic flow chart of a risk monitoring method according to an embodiment of the present invention.
Fig. 2 is a schematic structural diagram of a risk monitoring device according to an embodiment of the present invention.
Fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. The exemplary embodiments and descriptions of the present invention are provided to explain the present invention, but not to limit the present invention. It should be noted that the embodiments and features of the embodiments in the present application may be arbitrarily combined with each other without conflict.
Fig. 1 is a schematic flow chart of a risk monitoring method according to an embodiment of the present invention, and as shown in fig. 1, the risk monitoring method according to the embodiment of the present invention includes:
step S1: inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data.
Step S2: and taking the output result of the preset risk monitoring model as a risk monitoring result.
In the step S1, the device inputs the fusion data into a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data. The apparatus may be a computer device for executing the method, and it should be noted that the client, product and enterprise-related data related to the embodiment of the present invention are authorized by the user.
The customer information profile data is knowledge profile data constructed from customer information, which may include:
basic information of the client, including: the client can check the financial asset scale at the end of the month, the client identity networking check certificate, the client credit rating and the like.
The basic information of family members associated with the client comprises: family members' last month financial asset size, networking audit certification, credit rating, etc.
Customer holdings enterprise information, including: the number of the held enterprises, the affiliated industries of the held enterprises, the held proportion, the held amount and the like.
The product information map data is knowledge map data constructed according to product information, and the product information may include:
trusting product bottom investment information, including: the configuration proportion of the bottom assets, the names of investment products, product codes, product types, the occupancy ratio and the like.
Trust product shelf life operation index data, including: unit net worth, historical net worth, and interval profitability, etc.
The disclosure report information returned by the trust company comprises: the contents of the disclosure report text and the date of the return of the disclosure report, etc.
The enterprise information map data is knowledge map data constructed according to enterprise information, and the enterprise information can comprise enterprise news and the like.
The risk monitoring map data corresponding to each information map data may be understood as map data for risk monitoring of each information map data.
And the client risk monitoring map data corresponding to the client information map data is used for carrying out risk monitoring on the client information map data.
And the product risk monitoring map data corresponding to the product information map data is used for carrying out risk monitoring on the product information map data.
And enterprise risk monitoring map data corresponding to the enterprise information map data is used for carrying out risk monitoring on the enterprise information map data.
Fusing sample data can be understood as the following data which is selected in advance and can be used as a neural network model for training:
customer dimension integrated knowledge graph data, product dimension integrated knowledge graph data, and enterprise dimension integrated knowledge graph data.
Neural Networks (NN) are complex network systems formed by a large number of simple processing units (called neurons) widely interconnected, reflect many basic features of human brain functions, and are highly complex nonlinear dynamical learning systems. The neural network has the capabilities of large-scale parallel, distributed storage and processing, self-organization, self-adaptation and self-learning, and is particularly suitable for processing inaccurate and fuzzy information processing problems which need to consider many factors and conditions simultaneously.
The training sample data are client dimension comprehensive knowledge graph data C1, product dimension comprehensive knowledge graph data C2 and enterprise dimension comprehensive knowledge graph data C3 which are constructed by stock data, and the training sample data comprise risk data, risk edge data and risk-free data. The training method is a conventional training method in the field and is not described in detail.
In the above step S2, the apparatus takes the output result of the preset risk monitoring model as the risk monitoring result.
According to the risk monitoring method provided by the embodiment of the invention, the fusion data is input to the preset risk monitoring model, and the output result of the preset risk monitoring model is used as the risk monitoring result, so that the labor cost can be reduced, the efficiency can be improved, the defect of limited data dimension can be overcome, and the risk early warning can be comprehensively carried out according to the service characteristics.
Further, the neural network model is obtained by combining the BRNN and the LSTM. The BRNN Bidirectional recurrent neural network (Bidirectional RNN) mainly solves the problem that elements of a front sequence cannot sense output of a rear sequence.
Long-short term memory (LSTM) is a special RNN, mainly to solve the problems of gradient extinction and gradient explosion during Long sequence training.
The relevance among data and the timeliness of the data are comprehensively considered, the model adopts a Bidirectional Recurrent Neural Network (BRNN), and meanwhile, the fact that the recurrent neural network stores context information in a limited mode and loses part of information with earlier time sequence time is considered, so that the network layer LSTM is added in the model for long and short time.
In summary, considering the context association, timeliness and information storage, the neural network model has five layers, including an Input Layer, a BRNN-Forward Layer, a BRNN-Backward Layer, an LSTM and an Output Layer.
The risk monitoring method provided by the embodiment of the invention can overcome the defect of a single model, thereby improving the model operation efficiency and the accuracy of the model output result.
Further, the risk monitoring results correspond to a customer dimension, a product dimension, and an enterprise dimension, respectively; correspondingly, after the step of taking the output result of the preset risk monitoring model as the risk monitoring result, the risk monitoring method further includes:
and if at least one of the client dimension risk monitoring result, the product dimension risk monitoring result and the enterprise dimension risk monitoring result has a risk, generating a risk early warning message corresponding to the at least one risk monitoring result having the risk. For example, if the product dimension risk monitoring result has a risk, a risk early warning message for the product dimension is generated.
The risk monitoring method provided by the embodiment of the invention can be used for acquiring the risk monitoring result with risk more pertinently and more timely.
Further, acquiring the fused data comprises:
respectively constructing customer information map data, product information map data and enterprise information map data; the customer information map data is recorded as A1, and the construction is specifically described as follows:
the basic information of the client, the basic information data of family members related to the client and the client stock enterprise information are used as structured and semi-structured data, and entities and attributes are extracted through a knowledge graph extraction technology to be used as nodes of a knowledge graph, wherein the nodes include but are not limited to (client name, client age, client credit rating, client asset scale, client stock enterprise number, client stock enterprise proportion and the like), and the relationships are used as edges of the knowledge graph to construct client information graph data A1. For example, < client A, holding enterprise, number of holding enterprises >, < client A, sister, family member B >, < family member B, credit rating > may constitute a knowledge-graph triple.
The steps of constructing the product information map data a2 and the enterprise information map data A3 may refer to the description of the steps of constructing the customer information map data a1, and will not be described in detail.
Respectively constructing customer risk monitoring map data, product risk monitoring map data and enterprise risk monitoring map data according to risk monitoring threshold values of data items respectively corresponding to the customer information map data, the product information map data and the enterprise information map data and corresponding data items; the customer risk monitoring map data is recorded as B1, and the construction is specifically described as follows:
obtaining data items corresponding to customer information profile data, which may be understood as data items for risk monitoring, may include:
asset fluctuation data, share fluctuation data, etc., the corresponding risk monitoring thresholds may be set at 30% and 10%, respectively.
The data items corresponding to the customer information map data may also be data unrelated to a risk monitoring threshold, such as a customer identification status result, for example, an abnormal status of a networking check result of the customer or an associated customer, or a bad record of a customer credit investigation result, and then the result is brought into a region of interest.
If the customer's financial asset size drops by more than 30% at the end of the month, then the data item of the asset fluctuation data is included in the area of interest.
If the stock holding proportion of the client stock holding enterprise is reduced by more than 10%, the data item of the stock fluctuation data is included in the attention area.
Constructing customer risk monitoring map data B1, for example, including:
and (4) forming knowledge graph triples by < asset fluctuation, threshold, 30% >, < threshold, 30% reduction, attention area >, < customer credit investigation, bad record and attention area >.
Further, acquiring the enterprise risk monitoring map data comprises:
acquiring enterprise information, identifying the text content of the enterprise information, and respectively matching the text content with words in a preset positive public opinion library and a preset negative public opinion library; the text content recognition technology is a mature technology in the field and is not described in detail. The preset positive public opinion library is a preset positive word corpus, and the preset negative public opinion library is a preset negative word corpus.
Determining the type of the text content according to the matching result; the type of the text content comprises positive content or negative content; if the matching result is a front word corpus in a preset front public opinion library, the text content is the front content; if the matching result is a negative word corpus in the preset negative public opinion library, the text content is negative content.
Traversing all the text contents, and acquiring the number of the text contents respectively corresponding to the positive content or the negative content according to the type of the text contents of each text content; examples are as follows:
all the character contents are 5 and are respectively marked as a to e, wherein a to b are positive contents, and c to e are negative contents; the number of contents corresponding to positive contents is 2 and the number of contents corresponding to negative contents is 3.
If the number of the text contents corresponding to the positive contents is less than the number of the text contents corresponding to the negative contents, determining that the enterprise information is negative enterprise information; referring to the above example, the business information is negative business information.
And acquiring enterprise interaction information corresponding to the negative enterprise information, and if the statistical value of at least one item of interaction index data in the enterprise interaction information is larger than a preset statistical data threshold value, taking the negative enterprise information as the enterprise risk monitoring map data. The enterprise interaction information can comprise interaction index data such as the number of praise, the number of comments, the number of forwarding and the like aiming at the enterprise information.
Taking the praise number as an example, if the corresponding preset statistical data threshold is set to 5000 times, if the praise number is greater than 5000 times, the negative enterprise information is used as the enterprise risk monitoring map data.
Negative enterprise information can be specifically placed in the region of interest as data content in the construction of enterprise risk monitoring map data.
The following may also be used:
dividing the positive content into four levels A-D according to the sequence of the positive influence degrees from high to low, dividing the negative content into four levels E-H according to the sequence of the negative influence degrees from low to high, classifying the negative enterprise information meeting the preset statistical data threshold value condition into a G level or an H level, putting the G level or the H level negative enterprise information into an attention area, and taking the attention area as the data content in the enterprise risk monitoring map data.
According to the risk monitoring method provided by the embodiment of the invention, the accuracy of the risk monitoring result can be further improved by reasonably acquiring the enterprise risk monitoring map data.
Further, obtaining the product risk monitoring profile data comprises:
investment products are classified into rights and interests, solid income, commodities, financial derivatives, mixed products and the like according to types, the proportion of various investment products and the running fluctuation rate of the products are obtained, and corresponding risk monitoring thresholds can be respectively set to be 30% and 20%. The calculation formula of the product operation fluctuation rate is mature technology in the field and is not described any further.
If the proportion of investment product exceeds 30%, the data item is taken into account in the region of interest.
If the product run fluctuation rate exceeds 10%, the data item is included in the region of interest.
The data items corresponding to the product information map data may also be data unrelated to risk monitoring thresholds, such as:
and (4) periodically crawling related disclosure reports or financial statements through preset enterprise official network addresses. And identifying various financial indexes in the report or report form.
And constructing a multivariate model, and predicting the financial failure possibility of the enterprise by using a Z scoring method. The indexes of the repayment capacity, the profit capacity and the operation capacity of the enterprise in each period can be reflected through five variables, and the probability of enterprise financial failure or bankruptcy can be comprehensively analyzed and predicted.
The enterprise repayment ability index X1 is (total operating capital/total assets) × 100, and X4 is (total market value of common stock and prior stock/total account value of liability) × 100.
The enterprise profit margin X2 ═ residual profit/total asset amount X100, and X3 ═ profit/total asset amount X100 before tax interest.
Enterprise operating capability X5 is sales revenue/total amount of assets.
Z is the discrimination function value, Z is W1 × X1+ W2 × X2+ W3 × X3+ W4 × X4+ W5 × X5, and when the Z value is lower than the discrimination threshold, Z is placed in the region of interest. The judgment threshold value can be set independently according to the actual situation, and can be selected to be 1.81.
W1 to W5 are weights corresponding to X1 to X5, and may be set independently according to actual conditions, and may be selected from 0.012, 0.014, 0.033, 0.006, and 0.999.
By reasonably acquiring the product risk monitoring map data, the accuracy of the risk monitoring result can be further improved.
The steps of constructing the product risk monitoring map data B2 and the enterprise risk monitoring map data B3 may refer to the description of the construction of the client risk monitoring map data B1, and are not described in detail.
And fusing the customer information map data, the customer risk monitoring map data, the product information map data, the product risk monitoring map data, the enterprise information map data and the enterprise risk monitoring map data respectively to obtain the customer dimension comprehensive knowledge map data, the product dimension comprehensive knowledge map data and the enterprise dimension comprehensive knowledge map data. The customer dimension integrated knowledge graph data is recorded as C1, and the specific construction is as follows:
matching the entities of the customer information profile data a1 and the customer risk monitoring profile data B1, for example: and (3) directing the fluctuation of the client assets in the client information map data A1 and the assets in the client risk monitoring map data B1 to the same entity, and carrying out knowledge merging by utilizing an entity link technology to realize the fusion of the client information map data A1 and the client risk monitoring map data B1 so as to obtain client dimension comprehensive knowledge map data C1. The steps of constructing the product dimension integrated knowledge graph data C2 and the enterprise dimension integrated knowledge graph data C3 may refer to the description of constructing the customer dimension integrated knowledge graph data C1, and are not described again.
The risk monitoring method provided by the embodiment of the invention is beneficial to accurately and comprehensively monitoring the risk by comprehensively fusing the data into the customer dimension comprehensive knowledge graph data C1, the product dimension comprehensive knowledge graph data C2 and the enterprise dimension comprehensive knowledge graph data C3.
Further, after the step of obtaining the fused data, the risk monitoring method further comprises:
and calculating the node weights in the customer dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data by using a utilization centrality calculation method.
Because each entity in the knowledge graph and the connection mode before the entity are different, including one-to-many, many-to-many and many-to-one modes, the importance degree of each node is also different. The degree of a node refers to the associated edge of the node, and the more the associated edges are, the greater the weight of the node is.
For example: the holdings business entity nodes, the customer entities in the customer information graph data a1, and the business share entities and business share threshold entities in the customer risk monitoring graph data B1 should be given greater weight than the customer entities with only related age.
The centrality calculation method is a mature method in the field and is not described in detail.
According to the risk monitoring method provided by the embodiment of the invention, the association relationship compactness among the entities is reflected by accurately calculating the node weight, so that the accuracy of the output result of the preset risk monitoring model is improved.
Further, after the step of calculating the node weights in the dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data of the customer by the utilization centrality calculating method, the risk monitoring method further comprises:
giving the node weight to entities in the dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data, and giving the node weight to the dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph dataVectorizing and representing the synthetic knowledge graph data; that is, the knowledge graph spectrum is subjected to vector representation, and entities and relationships in the knowledge graph need to be mapped into space vectors through different mapping matrixes for representation. By triplets<Customer A, holding enterprise, number of holding enterprises>For example, a head entity client A is trained through a word2vec model to obtain a vector h, a relation/attribute supporting enterprise vector is r, a tail entity supporting enterprise quantity vector is t, and the entity and the relation vector can be mapped into a space vector h through a mapping function M respectivelyT＝MrhX h and tT＝Mrh×t。
And performing operation processing on the client dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data which are represented in the quantitative manner by using a MainfoldE algorithm. The MainfoldE algorithm is a mature algorithm in the field, and maps the relation r to a flow body, namely h + r is a hypersphere with the center radius of r, and is not an accurate point approximate to h + r, so that the method is more beneficial to processing a knowledge graph with complex relation.
The risk monitoring method provided by the embodiment of the invention is more beneficial to processing the knowledge graph with complex relation.
Further, the risk monitoring method further comprises:
inputting the fusion data which is subjected to the operation processing and is expressed in a vectorization mode to the preset risk monitoring model, and continuing to execute the subsequent steps. That is, in this step, the fusion data is replaced with the fusion data input with vectorization representation after operation, and other descriptions are omitted.
It can be understood that the fusion data expressed by vectorization after operation processing is used as the input of the model, so that the operation efficiency of the preset risk monitoring model can be improved.
The risk monitoring method provided by the embodiment of the invention can further improve the operation efficiency of the preset risk monitoring model.
It should be noted that the risk monitoring method provided by the embodiment of the present invention may be used in the financial field, and may also be used in any technical field except the financial field.
Fig. 2 is a schematic structural diagram of a risk monitoring device according to an embodiment of the present invention, and as shown in fig. 2, the risk monitoring device according to the embodiment of the present invention includes an input unit 201 and a monitoring unit 202, where:
the input unit 201 is used for inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data; the monitoring unit 202 is configured to use an output result of the preset risk monitoring model as a risk monitoring result.
Specifically, an input unit 201 in the device is used for inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data; the monitoring unit 202 is configured to use an output result of the preset risk monitoring model as a risk monitoring result.
According to the risk monitoring device provided by the embodiment of the invention, the fusion data is input to the preset risk monitoring model, and the output result of the preset risk monitoring model is used as the risk monitoring result, so that the labor cost can be reduced, the efficiency can be improved, the defect of limited data dimension can be overcome, and the risk early warning can be comprehensively carried out according to the service characteristics.
The embodiment of the risk monitoring device provided in the embodiment of the present invention may be specifically configured to execute the processing flows of the above method embodiments, and the functions of the embodiment are not described herein again, and refer to the detailed description of the above method embodiments.
Fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 3, the electronic device includes: a processor (processor)301, a memory (memory)302, and a bus 303;
the processor 301 and the memory 302 complete communication with each other through a bus 303;
the processor 301 is configured to call program instructions in the memory 302 to perform the methods provided by the above-mentioned method embodiments, including:
inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data;
and taking the output result of the preset risk monitoring model as a risk monitoring result.
The present embodiment discloses a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the method provided by the above-mentioned method embodiments, for example, comprising:
inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data;
and taking the output result of the preset risk monitoring model as a risk monitoring result.
The present embodiment provides a computer-readable storage medium, which stores a computer program, where the computer program causes the computer to execute the method provided by the above method embodiments, for example, the method includes:
inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data;
and taking the output result of the preset risk monitoring model as a risk monitoring result.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In the description herein, reference to the description of the terms "one embodiment," "a particular embodiment," "some embodiments," "for example," "an example," "a particular example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
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 only 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 (11)
1. A method of risk monitoring, comprising:
inputting fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data;
and taking the output result of the preset risk monitoring model as a risk monitoring result.
2. The risk monitoring method of claim 1, wherein the neural network model is obtained by combining BRNN and LSTM.
3. The risk monitoring method of claim 1, wherein the risk monitoring results correspond to a customer dimension, a product dimension, and an enterprise dimension, respectively; correspondingly, after the step of taking the output result of the preset risk monitoring model as the risk monitoring result, the risk monitoring method further includes:
and if at least one of the client dimension risk monitoring result, the product dimension risk monitoring result and the enterprise dimension risk monitoring result has a risk, generating a risk early warning message corresponding to the at least one risk monitoring result having the risk.
4. The risk monitoring method of claim 1, wherein obtaining the fused data comprises:
respectively constructing customer information map data, product information map data and enterprise information map data;
respectively constructing customer risk monitoring map data, product risk monitoring map data and enterprise risk monitoring map data according to risk monitoring threshold values of data items respectively corresponding to the customer information map data, the product information map data and the enterprise information map data and corresponding data items;
and fusing the customer information map data, the customer risk monitoring map data, the product information map data, the product risk monitoring map data, the enterprise information map data and the enterprise risk monitoring map data respectively to obtain the customer dimension comprehensive knowledge map data, the product dimension comprehensive knowledge map data and the enterprise dimension comprehensive knowledge map data.
5. The risk monitoring method of claim 4, wherein after the step of acquiring the fused data, the risk monitoring method further comprises:
and calculating the node weights in the customer dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data by using a utilization centrality calculation method.
6. The risk monitoring method of claim 5, wherein after the step of the centrality of utilization calculation method calculating node weights in the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data, the risk monitoring method further comprises:
endowing the node weight to entities in the dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data, and performing vectorization expression on the dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data which are endowed with the node weight;
and performing operation processing on the client dimension comprehensive knowledge graph data, the product dimension comprehensive knowledge graph data and the enterprise dimension comprehensive knowledge graph data which are represented in the quantitative manner by using a MainfoldE algorithm.
7. The risk monitoring method of claim 6, further comprising:
inputting the fusion data which is subjected to the operation processing and is expressed in a vectorization mode to the preset risk monitoring model, and continuing to execute the subsequent steps.
8. The risk monitoring method of claim 4, wherein obtaining the enterprise risk monitoring profile data comprises:
acquiring enterprise information, identifying the text content of the enterprise information, and respectively matching the text content with words in a preset positive public opinion library and a preset negative public opinion library;
determining the type of the text content according to the matching result; the type of the text content comprises positive content or negative content;
traversing all the text contents, and acquiring the number of the text contents respectively corresponding to the positive content or the negative content according to the type of the text contents of each text content;
if the number of the text contents corresponding to the positive contents is less than the number of the text contents corresponding to the negative contents, determining that the enterprise information is negative enterprise information;
and acquiring enterprise interaction information corresponding to the negative enterprise information, and if the statistical value of at least one item of interaction index data in the enterprise interaction information is larger than a preset statistical data threshold value, taking the negative enterprise information as the enterprise risk monitoring map data.
9. A risk monitoring device, comprising:
the input unit is used for inputting the fusion data to a preset risk monitoring model; the preset risk monitoring model is obtained by training a neural network model according to the fusion sample data; the fusion data is fused with customer dimension comprehensive knowledge graph data, product dimension comprehensive knowledge graph data and enterprise dimension comprehensive knowledge graph data; the customer dimension integrated knowledge graph data, the product dimension integrated knowledge graph data and the enterprise dimension integrated knowledge graph data comprise information graph data corresponding to each customer dimension integrated knowledge graph data, and risk monitoring graph data corresponding to each information graph data;
and the monitoring unit is used for taking the output result of the preset risk monitoring model as a risk monitoring result.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method of any of claims 1 to 8 are implemented when the computer program is executed by the processor.
11. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210012741.6A CN114358607A (en) | 2022-01-06 | 2022-01-06 | Risk monitoring method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210012741.6A CN114358607A (en) | 2022-01-06 | 2022-01-06 | Risk monitoring method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114358607A true CN114358607A (en) | 2022-04-15 |
Family
ID=81106309
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210012741.6A Pending CN114358607A (en) | 2022-01-06 | 2022-01-06 | Risk monitoring method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114358607A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108596439A (en) * | 2018-03-29 | 2018-09-28 | 北京中兴通网络科技股份有限公司 | A kind of the business risk prediction technique and system of knowledge based collection of illustrative plates |
| CN112597775A (en) * | 2020-12-16 | 2021-04-02 | 中国工商银行股份有限公司 | Credit risk prediction method and device |
| CN112598489A (en) * | 2020-12-14 | 2021-04-02 | 深圳市快付通金融网络科技服务有限公司 | Risk monitoring method and system based on financial application |
| CN113822494A (en) * | 2021-10-19 | 2021-12-21 | 平安科技（深圳）有限公司 | Risk prediction method, device, equipment and storage medium |
-
2022
- 2022-01-06 CN CN202210012741.6A patent/CN114358607A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108596439A (en) * | 2018-03-29 | 2018-09-28 | 北京中兴通网络科技股份有限公司 | A kind of the business risk prediction technique and system of knowledge based collection of illustrative plates |
| CN112598489A (en) * | 2020-12-14 | 2021-04-02 | 深圳市快付通金融网络科技服务有限公司 | Risk monitoring method and system based on financial application |
| CN112597775A (en) * | 2020-12-16 | 2021-04-02 | 中国工商银行股份有限公司 | Credit risk prediction method and device |
| CN113822494A (en) * | 2021-10-19 | 2021-12-21 | 平安科技（深圳）有限公司 | Risk prediction method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220075670A1 (en) | Systems and methods for replacing sensitive data | |
| Wu | Supplier selection: A hybrid model using DEA, decision tree and neural network | |
| EP4236197A2 (en) | Micro-loan system | |
| US11526695B2 (en) | Evaluating impact of process automation on KPIs | |
| CN109949152A (en) | A kind of personal credit's violation correction method | |
| CN110163413A (en) | Enterprise supervision and method for early warning, device, computer equipment and readable storage medium storing program for executing | |
| Wei | [Retracted] A Method of Enterprise Financial Risk Analysis and Early Warning Based on Decision Tree Model | |
| CN112241805A (en) | Defect prediction using historical inspection data | |
| Castañeda et al. | The resilience of public policies in economic development | |
| CN114638498A (en) | ESG evaluation method, ESG evaluation system, electronic equipment and storage equipment | |
| Wang et al. | Multiview Graph Learning for Small‐and Medium‐Sized Enterprises’ Credit Risk Assessment in Supply Chain Finance | |
| CN108711074B (en) | Service classification method, device, server and readable storage medium | |
| Gleue et al. | Decision support for the automotive industry: Forecasting residual values using artificial neural networks | |
| Camelia et al. | A Computational Grey Based Model for Companies Risk Forecasting. | |
| CN117114812A (en) | Financial product recommendation method and device for enterprises | |
| Liu et al. | Risk assessment and regulation algorithm for financial technology platforms in smart city | |
| Ocampo et al. | A sustainable manufacturing strategy from different strategic responses under uncertainty | |
| Lederer et al. | Data science for business analytics and business intelligence | |
| CN114358607A (en) | Risk monitoring method and device | |
| Li et al. | Construction of Credit Evaluation Index System for Two‐Stage Bayesian Discrimination: An Empirical Analysis of Small Chinese Enterprises | |
| US20220101232A1 (en) | Inferential analysis and reporting of contextual complaints data | |
| AlGosaibi et al. | Developing an Intelligent Framework for Improving the Quality of Service in the Government Organizations in the Kingdom of Saudi Arabia | |
| Sönmez et al. | Big data analytics and models | |
| Valan et al. | Data analytics for relative ranking of factors to optimise blood bank supply chain | |
| Corekcioglu et al. | Estimation of success of entrepreneurship projects with data mining |
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 |