CN113989019A

CN113989019A - Method, device, equipment and storage medium for identifying risks

Info

Publication number: CN113989019A
Application number: CN202111256742.7A
Authority: CN
Inventors: 石大文; 张艳停
Original assignee: Ping An Bank Co Ltd
Current assignee: Ping An Bank Co Ltd
Priority date: 2021-10-27
Filing date: 2021-10-27
Publication date: 2022-01-28

Abstract

The application is applicable to the technical field of artificial intelligence, and provides a method, a device, equipment and a storage medium for risk identification. The method comprises the following steps: collecting target data; extracting a plurality of entities, attribute information of each entity and relationship information among the entities from the acquired target data and constructing an initial risk knowledge graph; introducing multiple risk factors of a user into an initial risk knowledge graph to generate a risk knowledge graph; determining risk data of a target user based on a risk knowledge graph; determining a risk index of the target user according to the risk data; and carrying out risk early warning on the target enterprise according to the risk index. According to the scheme, the risk factors are introduced into the constructed risk knowledge graph, deeper and more accurate risk data can be mined from the risk knowledge graph according to the risk factors, the risk index is determined according to the risk data, risk early warning is timely provided for enterprises, financial fraud events can be effectively prevented, and economic and safety guarantee is provided for the enterprises.

Description

Method, device, equipment and storage medium for identifying risks

Technical Field

The application belongs to the technical field of artificial intelligence, and particularly relates to a risk identification method, device, equipment and storage medium.

Background

With the development of science and technology, wind control is of great importance to the financial industry. Through the analysis of mass data, risk prediction is carried out on the user, and the risk prediction result can be used as an indispensable part for evaluating the credit of the user in credit business of banks or other financial institutions.

At present, in the field of wind control of financial industry, an enterprise cannot accurately predict risks of users, and cannot accurately perform risk early warning for the enterprise in time, so that financial fraud events frequently occur, and economic losses are caused for financial enterprises.

Disclosure of Invention

In view of this, embodiments of the present application provide a method, an apparatus, a device, and a storage medium for risk identification, so as to solve the problem in the prior art that a risk of a user cannot be accurately predicted, and thus a risk early warning cannot be timely and accurately performed on an enterprise, which further causes frequent financial fraud events and economic loss for financial enterprises.

A first aspect of an embodiment of the present application provides a method for identifying a risk, where the method includes:

acquiring target data for constructing a risk knowledge graph;

extracting a plurality of entities, attribute information of each entity and relationship information among the entities from the acquired target data, wherein the entities comprise users, enterprises and merchants;

constructing an initial risk knowledge graph according to the extracted entities, the attribute information of the entities and the relationship information among the entities;

acquiring multiple risk factors of the user, importing the multiple risk factors of the user into the initial risk knowledge graph to generate the risk knowledge graph, wherein the multiple risk factors comprise a transaction mode, overdue records, credit of a transaction object and a risk merchant;

determining risk data of a target user based on the risk knowledge graph;

determining a risk index of the target user according to the risk data;

and carrying out risk early warning on the target enterprise according to the risk index.

Optionally, the determining a risk index of the target user according to the risk data includes:

acquiring a weight value and an evaluation rule corresponding to each risk factor respectively;

calculating a score corresponding to each risk factor according to the weight value and the evaluation rule respectively corresponding to each risk factor and the risk data;

and determining the risk index according to the corresponding score of each risk factor.

Optionally, the calculating a score corresponding to each of the risk factors according to the weight value and the evaluation rule respectively corresponding to each of the risk factors and the risk data includes:

for each risk factor, acquiring target risk data corresponding to the risk factor in the risk data;

evaluating the target risk data according to an evaluation rule corresponding to the risk factor to obtain an initial score corresponding to the target risk data;

and calculating the product of the weight value corresponding to the risk factor and the initial score, and determining the score corresponding to the risk factor based on the product.

Optionally, the performing risk early warning on the target enterprise according to the risk index includes:

determining a risk grade corresponding to the risk index;

and carrying out risk early warning of different levels on the target enterprise according to the risk level.

Optionally, before determining the risk index of the target user according to the risk data, the method further includes:

removing the overtop data in the risk data, wherein the overtop data comprises transaction data of the target user and white list merchants, and the white list merchants are merchants transacting with a plurality of users for more than preset times;

determining a risk index for the target user based on the risk data comprises:

and determining the risk index of the target user according to the risk data after the over point data is removed.

Optionally, after performing risk early warning on the target enterprise according to the risk index, the method further includes:

and generating a risk log according to the multiple risk factors, the target user and the risk data.

Optionally, before extracting a plurality of entities, attribute information of each entity, and relationship information between entities from the collected target data, the method further includes:

converting the format of the target data into a JSON format;

the extracting of the plurality of entities, the attribute information of each entity and the relationship information among the entities from the collected target data includes:

and extracting a plurality of entities, attribute information of each entity and relationship information among the entities from the target data in the JSON format.

A second aspect of an embodiment of the present application provides an apparatus for identifying risk, including:

the acquisition unit is used for acquiring target data for constructing a risk knowledge graph;

the extraction unit is used for extracting a plurality of entities, attribute information of each entity and relationship information among the entities from the acquired target data, wherein the entities comprise users, enterprises and merchants;

the construction unit is used for constructing an initial risk knowledge graph according to each extracted entity, the attribute information of each entity and the relationship information among the entities;

the acquisition unit is used for acquiring various risk factors of the user, importing the various risk factors of the user into the initial risk knowledge graph and generating the risk knowledge graph, wherein the various risk factors comprise a transaction mode, overdue records, credit of a transaction object and a risk merchant;

a first determining unit, configured to determine risk data of a target user based on the risk knowledge graph;

the second determining unit is used for determining the risk index of the target user according to the risk data;

and the early warning unit is used for carrying out risk early warning on the target enterprise according to the risk index.

A third aspect of the embodiments of the present application provides a device for identifying risk, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the method for identifying risk according to the first aspect.

A fourth aspect of embodiments of the present application provides a computer-readable storage medium, in which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of identifying risk as described in the first aspect above.

A fifth aspect of embodiments of the present application provides a computer program product, which when run on an apparatus, causes the apparatus to perform the steps of the method for identifying risk according to the first aspect.

The method, the device, the equipment and the storage medium for identifying the risks have the following beneficial effects:

acquiring target data for constructing a risk knowledge graph; extracting a plurality of entities, attribute information of each entity and relationship information among the entities from the acquired target data, wherein the entities comprise users, enterprises and merchants; constructing an initial risk knowledge graph according to each extracted entity, attribute information of each entity and relationship information among the entities; acquiring multiple risk factors of a user, and importing the multiple risk factors of the user into an initial risk knowledge graph to generate a risk knowledge graph, wherein the multiple risk factors comprise a transaction mode, overdue records, credit of a transaction object and a risk merchant; determining risk data of a target user based on a risk knowledge graph; determining a risk index of the target user according to the risk data; and carrying out risk early warning on the target enterprise according to the risk index. According to the scheme, multiple different risk factors are introduced when the risk knowledge graph is constructed, deeper and more accurate risk data can be mined from the risk knowledge graph according to the risk factors, and then the risk index is determined according to the risk data, so that risk early warning is provided for enterprises in time, financial fraud events can be effectively prevented, and economic and safety guarantee is provided for the enterprises. And according to the risk knowledge graph, risk data are determined, the data processing speed is increased, and the risk early warning efficiency of enterprises is further improved.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise.

FIG. 1 is a schematic flow chart diagram of a method of identifying risk provided by an exemplary embodiment of the present application;

FIG. 2 is a detailed flowchart of step S106 of a method for identifying risk according to yet another exemplary embodiment of the present application;

FIG. 3 is a detailed flowchart of step S107 of a method for identifying risk according to another exemplary embodiment of the present application;

FIG. 4 is a detailed flow diagram of a method of constructing a risk profile according to yet another exemplary embodiment of the present application;

FIG. 5 is a schematic diagram of an apparatus for identifying risk provided by an embodiment of the present application;

fig. 6 is a schematic diagram of an apparatus for identifying risk according to another embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

In the description of the embodiments of the present application, "/" means "or" unless otherwise specified, for example, a/B may mean a or B; "and/or" herein is merely an association describing an associated object, and means that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of the present application, "a plurality" means two or more than two.

In the following, the terms "first", "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present embodiment, "a plurality" means two or more unless otherwise specified.

The embodiment of the application can acquire and process related data based on an artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result.

With the development of science and technology, the financial industry and digital technology are rapidly developed and fused, and meanwhile, the risk of financial fraud is continuously expanded. Digital financial fraud gradually shows the characteristics of specialization, industrialization, concealment, scene and the like; compared with the traditional fraud, the digital financial fraud is organized and scaled, and the digital financial fraud is clear in work division, close in cooperation and in cooperative case, thereby forming a complete crime industry chain. For example, fraudulent credits from a user organization with poor credit often occur, resulting in the inability of the business to receive the payout.

In the face of complex big data, how to efficiently acquire valuable information from the large-scale data, how to quickly identify cheating groups and general cheating means and take measures in time is a huge challenge faced by the traditional technology.

At present, in the wind control field of the financial industry, the risk control capability of an enterprise is poor, the risk of a user cannot be accurately predicted based on big data, and risk early warning cannot be timely and accurately performed on the enterprise, so that financial fraud events frequently occur, and economic loss is caused for the financial enterprise.

In view of this, the embodiment of the present application provides a method for identifying a risk, which collects target data for constructing a risk knowledge graph; extracting a plurality of entities, attribute information of each entity and relationship information among the entities from the acquired target data, wherein the entities comprise users, enterprises and merchants; constructing an initial risk knowledge graph according to each extracted entity, attribute information of each entity and relationship information among the entities; acquiring multiple risk factors of a user, and importing the multiple risk factors of the user into an initial risk knowledge graph to generate a risk knowledge graph, wherein the multiple risk factors comprise a transaction mode, overdue records, credit of a transaction object and a risk merchant; determining risk data of a target user based on a risk knowledge graph; determining a risk index of the target user according to the risk data; and carrying out risk early warning on the target enterprise according to the risk index.

According to the scheme, multiple different risk factors are introduced when the risk knowledge graph is constructed, deeper and more accurate risk data can be mined from the risk knowledge graph according to the risk factors, and then the risk index is determined according to the risk data, so that risk early warning is provided for enterprises in time, financial fraud events can be effectively prevented, and economic and safety guarantee is provided for the enterprises. And according to the risk knowledge graph, risk data are determined, the data processing speed is increased, and the risk early warning efficiency of enterprises is further improved.

Referring to fig. 1, fig. 1 is a schematic flow chart of a method for identifying risk provided by an exemplary embodiment of the present application. The method for identifying risks provided by the present application is executed by a device, wherein the device includes, but is not limited to, a mobile terminal such as a smart phone, a tablet computer, a Personal Digital Assistant (PDA), a desktop computer, and may further include various types of servers. For example, the server may be an independent server, or may be a cloud service that provides basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a web service, cloud communication, a middleware service, a domain name service, a security service, a Content Delivery Network (CDN), and a big data and artificial intelligence platform.

In the embodiments of the present application, an execution subject is taken as an example for description.

The method of identifying risk as shown in fig. 1 may include: s101 to S107 are as follows:

s101: target data for constructing a risk profile is collected.

Illustratively, different target data can be collected according to different business scenarios of an enterprise, and different risk knowledge graphs can be constructed. For example, according to a loan service of a certain enterprise, data required for constructing a risk knowledge graph matching the service scenario is collected.

For example, in the context of a loan transaction, the target data collected may include: user data, merchant data, enterprise data, and the like.

The user data may include personal information of the user, and transaction records, transfer data, consumption data, payment data, asset data, etc. of the user, among others. For example, the personal information of the user includes the user's name, sex, age, identification number, contact address, home address, work situation, and the like. The transaction record of the user can comprise data generated when the user transacts with any merchant; the transfer data can comprise data generated when a user transfers with any object through any channel; consumption data may include data generated when consumed by a user; the repayment data may include all repayment data under the name of the user; asset data may include balance information, financial information, house information, vehicle information, payroll information, etc. of a user.

The merchant data may include water flow information of the merchant in a preset time period, data generated by transactions between the merchant and various users, and the like.

Enterprise data may include registered capital, financial information, various news opinions, legal decisions documents, business information, and the like.

For example, the target data for constructing the risk knowledge graph may be collected in the network, or may be collected in the database of each financial enterprise, which is only an exemplary illustration and is not limited herein.

S102: and extracting a plurality of entities, attribute information of each entity and relationship information among the entities from the collected target data, wherein the entities comprise users, enterprises and merchants.

In this example, knowledge extraction is performed on the collected target data. Specifically, entity extraction is carried out on the collected target data to obtain a plurality of entities; performing attribute extraction on the acquired target data to obtain attribute information corresponding to each entity; and extracting the relation of the collected target data to obtain the relation information among the entities.

For example, in the scenario of a loan transaction, the collected target data may include user data, merchant data, enterprise data, and the like. Each user, merchant, enterprise, various objects transacting with the user, etc. may be entities.

And taking the basic information of each entity as the attribute information of the entity. For example, the personal information of the user is the attribute information of the user.

And regarding the association relationship among the entities as relationship information among the entities. For example, a user has consumed at a certain merchant, the association relationship between the user and the merchant is the consumption transaction, and the consumption transaction can be used as the relationship information between the user and the merchant.

For another example, a user transacts a loan transaction in a financial enterprise, the association relationship between the user and the financial enterprise is the loan transaction, and the loan transaction can be regarded as the relationship information between the user and the financial enterprise. The description is given for illustrative purposes only and is not intended to be limiting.

Optionally, an extraction rule may be customized, and according to the extraction rule, a plurality of entities, attribute information of each entity, and relationship information between the entities are extracted from the collected target data. The method can also be used for training an extraction model through a traditional machine learning algorithm, and processing the acquired target data through the extraction model to obtain a plurality of entities, attribute information of each entity and relationship information among the entities. The description is given for illustrative purposes only and is not intended to be limiting.

Optionally, in a possible implementation manner, before performing S102, the method may further include: and uniformly converting the format of the collected target data into a lightweight data interchange format (JSON). And then extracting a plurality of entities, attribute information of each entity and relationship information among the entities from the target data in the JSON format.

The format of the target data is uniformly converted into the JSON format, so that the target data is easy to read and write, the machine analysis is facilitated, the data transmission efficiency is improved, and the speed of constructing the risk knowledge graph is also improved.

S103: and constructing an initial risk knowledge graph according to each extracted entity, the attribute information of each entity and the relationship information among the entities.

The initial risk profile is essentially a semantic network, a graph-based data structure, consisting of nodes and edges. In the initial risk knowledge graph, each node represents an "entity" existing in the real world, and each edge is a "relationship" between entities.

In the initial risk knowledge graph, each entity has a globally unique identification information (e.g., ID) for identification and indexing, and each entity may have several different attribute information (Property) for characterizing the entity. Edges (edges) are used to describe the relationship between two entities. In general, the initial risk knowledge graph is a relationship network obtained by connecting together each extracted entity, attribute information of each entity, and relationship information between entities.

Illustratively, entity connection is performed on the extracted multiple entities, attribute information of each entity and relationship information among the entities, that is, the information is integrated to form a knowledge graph of the relationship among the entities. And adding the integrated knowledge graph information into a structured knowledge base, and performing knowledge graph demonstration by using knowledge reasoning of the knowledge base so as to obtain an initial risk knowledge graph. In the initial risk knowledge-graph, entities are stored in the form of [ entity-relationship-entity ]. For each entity, attribute information corresponding to each entity is stored in a mode of entity-attribute-value.

Exemplarily, the extracted multiple entities, the attribute information of each entity, and the relationship information between the entities are stored in a triple storage format. For example, the extracted multiple entities, attribute information of each entity, and relationship information between the entities may be stored in a matrix form, so as to obtain an initial risk knowledge graph.

For example, there are two entities that are users: zhang III and Enterprise: and XX bank, transacting the credit card in XX bank by Zhang III in the initial risk knowledge graph. The two entities are stored in the form of [ three (entities) -transacting credit card (relationship) -XX bank (entity) ]. For the user Zhang three, the entity stores the corresponding attribute information in the mode of entity-attribute-value. For example, the attribute information of Zhang III is sex male, and may be stored as [ Zhang III-sex-male ]. The description is given for illustrative purposes only and is not intended to be limiting.

It should be noted that the initial risk knowledge-graph and the final risk knowledge-graph are identical in representation form, that is, in the risk knowledge-graph, the entities are also stored in the form of [ entity-relationship-entity ]. For each entity, attribute information corresponding to each entity is stored in a mode of entity-attribute-value. In contrast, the risk knowledge graph incorporates multiple risk factors for the user, while the initial risk knowledge graph does not store multiple risk factors for the user. In the risk knowledge graph, a variety of risk factors for a user may be stored in the attribute information of each user.

The method for constructing the initial risk knowledge graph simplifies data from complexity and improves the efficiency of constructing the initial risk knowledge graph.

S104: and acquiring various risk factors of the user, importing the various risk factors of the user into the initial risk knowledge graph, and generating the risk knowledge graph, wherein the various risk factors comprise a transaction mode, overdue records, the credit of a transaction object and a risk merchant.

Illustratively, the risk factors may include: any combination of transaction mode, number of transacted people, transaction amount, overdue records, credit of transacted objects, number of overdue records in transacted objects, risk merchants and the like.

For example, different risk factors may be preset according to different businesses of different enterprises. For example, the risk factors may include: any combination or any one of transaction mode, number of transacted people, transaction amount, overdue records, credit of transacted objects, number of overdue records in transacted objects, risk merchants and the like.

Illustratively, the transaction means may in turn include transfers (e.g., payroll transfers, wechat transfers, internet banking transfers, direct bank remittance, etc.), consumer payments (e.g., network consumer payments, physical store consumer payments, etc.), and the like.

When the transaction mode is transfer, the number of transaction persons can include the number of transfer persons, namely the number of transfer transactions with the user in a preset time period; when the transaction mode is consumption payment, the number of transaction persons may include the number of persons having a consumption payment transaction with the user within a preset time period.

When the transaction mode is transfer, the transaction amount can comprise transfer amount; when the transaction mode is payment for consumption, the transaction amount may include a payment amount.

The overdue records may include records of overdue payments made by the user on various platforms (e.g., various loanable platforms).

The credit for the transaction object may include the credit for an object with which a transfer transaction occurred, the credit for an object with which a consumption payment transaction occurred, etc.

The number of people with overdue records in the traded object may include the number of people with overdue records in all objects that have a money transfer transaction with the user.

The risky merchant may include a merchant that provides a cash register service (e.g., a cash register, a credit card cash register, etc.) for the user.

And acquiring various risk factors of the user according to different services. For example, in the scenario of a loan transaction, the various risk factors of the user that are obtained may include: the number of transfer accounts, overdue records, the credit of transfer objects, the number of persons with overdue records in all transfer objects, risk merchants and the like.

Illustratively, a plurality of risk factors for the user are imported into the initial risk knowledge graph. And acquiring risk data corresponding to various risk factors according to different risk factors while importing the risk factors, and storing the risk data in the attribute information of each user or storing the risk data in the relationship information of each user.

For example, when the risk factors are imported, the risk factors may be stored in the node corresponding to each user entity.

And the initial risk knowledge graph after the plurality of risk factors of the user are imported is the risk knowledge graph.

In the embodiment, the risk factors are introduced into the construction of the risk knowledge graph, and various risk factors are considered in multiple directions, so that the subsequently generated risk index is more representative and more accurate, and the early warning can be timely and accurately performed on the target enterprise.

S105: based on the risk knowledge graph, risk data of the target user is determined.

Illustratively, the risk knowledge graph is a knowledge base with a graph structure, and knowledge units with different sources, different types and different structures can be linked and associated into a graph, so that a wider and deeper knowledge system is provided for each financial enterprise and is continuously expanded.

The risk data refers to data generated after the user performs an operation matched with the risk factor. For example, the risk factor is the number of transfer accounts, the risk data is the transfer record of the target user in a preset time period, and the number of transfer accounts is obtained through statistics according to the transfer record. For another example, the risk factor is overdue record, and the risk data is overdue record of the target user in the business of credit card, loan, etc. transacted by the financial enterprise. The description is given for illustrative purposes only and is not intended to be limiting.

Illustratively, the constructed risk knowledge graph can be stored in a local database in advance, and when risk prediction is needed, the server sends an acquisition request to the local terminal, wherein the acquisition request is used for acquiring the risk knowledge graph. And the local terminal sends the risk knowledge graph to the server, and the server acquires the risk knowledge graph.

Or the constructed risk knowledge graph can be stored in the server in advance, and the server directly calls the risk knowledge graph when risk identification is needed.

The risk knowledge graph comprises a plurality of entities, attribute information of each entity, relationship information among the entities and risk factors.

Illustratively, globally unique identification information is set for each entity when constructing the risk knowledge graph. When a target user is to be queried, acquiring identification information corresponding to the target user, and searching a target entity corresponding to the target user in the risk knowledge graph according to the identification information of the target user. And inquiring the risk data related to the target entity in the risk knowledge graph according to the risk factor.

The risk factors are introduced when the risk knowledge graph is constructed. When the risk factors are imported, the risk factors are stored in the node corresponding to each user entity. When a target user is to be queried, the target entity is searched according to the identification information of the target user, which risk factors exist in the target entity is searched, and the risk data corresponding to the risk factors of the target entity is queried in the risk knowledge graph according to each risk factor. When the risk factors are various, various risk data are inquired. Wherein each risk data corresponds to a risk factor.

S106: and determining the risk index of the target user according to the risk data.

For example, when there are a plurality of risk factors and a plurality of risk data corresponding to the target user, determining a score corresponding to each risk data, calculating the sum of all scores, and taking the sum of all scores as the risk index of the target user. The description is given for illustrative purposes only and is not intended to be limiting.

Optionally, when there is only one risk factor, there is only one risk data corresponding to the target user, and the score determined according to the risk data is the risk index of the target user.

Optionally, in a possible implementation manner, before performing S106, the method may further include: and eliminating the overtop data in the risk data, wherein the overtop data comprises transaction data of the target user and white list merchants, and the white list merchants are merchants having transactions with a plurality of users for more than preset times. And determining the risk index of the target user according to the risk data after the over-point data is removed.

Illustratively, the over-point data is not considered in the risk index calculation for the target user. For example, a certain catering merchant pays by means of money transfer, and the number of money transfers with the catering merchant is recorded for more than a preset number of times within a preset time period, but since the catering merchant has no problem in the transaction between the user and the catering merchant, when the catering merchant is used as a risk factor to calculate the risk index of the user, the credit evaluation of the user is not helped, and the final evaluation result is influenced. Thus, the over-point data can be predetermined and added to the white list.

For example, when inquiring merchants with transaction times exceeding a certain number in a preset time period, adding an entity corresponding to the merchant into a white list, and inquiring risk data of a target user in a risk knowledge graph, even if the target user has traded with the merchant, the transaction of the target user and the merchant is not considered due to the fact that the merchant is added into the white list.

By the implementation mode, interference caused by the over-point data is eliminated, so that the accuracy of the risk index is improved, and the risk early warning is more accurate.

S107: and carrying out risk early warning on the target enterprise according to the risk index.

Exemplarily, the risk level corresponding to the risk index can be determined, and different forms of risk early warning are performed on the target enterprise according to different risk levels.

Optionally, in a possible implementation manner, risk early warning is directly performed on the target enterprise according to the risk index without determining the risk level corresponding to the risk index.

For example, in the scenario of loan transaction, when a user needs to transact loan transaction or credit card transaction, the financial enterprise will automatically check whether the user can transact loan transaction or credit card transaction through its own system. For example, the loan transaction or the credit card transaction may be wrongly handled by the risky user, because the user generally submits personal information by auditing, so that the audited information is not comprehensive, and the auditing result is not accurate. Since this user is actually an inauguration user, difficulty occurs in making a payment later, causing economic loss (for example, bad account) of the financial enterprise. Colloquially, fraudulent crediting may occur.

Illustratively, the risk pre-warning manner may include: message reminders, voice reminders, and the like. For example, a reminder message may be generated to remind the enterprise staff to change the automatic review of the target user into a manual review, or to reject the service request of the target user, etc. Or the system automatically changes the information audit of the target user to manual audit, and reminds enterprise staff to audit the information and business of the target user. The description is given for illustrative purposes only and is not intended to be limiting.

According to the scheme, the risk factors are introduced into the constructed risk knowledge graph, deeper and more accurate risk data can be mined from the risk knowledge graph according to the risk factors, and then the risk index is determined according to the risk data, so that risk early warning is provided for enterprises in time, financial fraud events can be effectively prevented from occurring, and economic and safety guarantee is provided for the enterprises.

Referring to fig. 2, fig. 2 is a flowchart illustrating a step S106 of a method for identifying risk according to another exemplary embodiment of the present application; optionally, in some possible implementations of the present application, the S106 may include S1061 to S1063, which are as follows:

s1061: and acquiring a weight value and an evaluation rule corresponding to each risk factor respectively.

For example, a different weight value may be set for each risk factor in advance, and an evaluation rule may be set for each risk factor and stored in the database. In implementation, the database may be queried for the weight value and evaluation rule corresponding to each risk factor.

For example, the risk factor is the number of transfer accounts, transfer records of a target user in a preset time period are inquired in the risk knowledge graph, and the number of transfer accounts is counted according to the transfer records, namely, the number of transfer accounts given by the target user in the preset time period is counted. The evaluation rule corresponding to the risk factor may include: whether the number of the transfer accounts exceeds a first preset number within a preset time period; whether the number of the transfer accounts exceeds a second preset number within a preset time period; whether the number of transfers exceeds a third preset number of transfers within a preset time period. Wherein the first preset number of people is less than the second preset number of people, and the second preset number of people is less than the third preset number of people. The description is given for illustrative purposes only and is not intended to be limiting.

For another example, the risk factor is an overdue record, and the overdue record of the target user is queried in the risk knowledge graph. For example, whether the target user has overdue records in the credit card, loan and other businesses transacted by other financial enterprises within a preset time period is queried. The number of overdue times of the target user may also be queried. The evaluation rule corresponding to the risk factor may include: whether overdue records exist or not within a preset time period; and whether the overdue times exceed a preset overdue threshold value within a preset time period. The description is given for illustrative purposes only and is not intended to be limiting.

For another example, the risk factor is credit of the transfer object, the transfer object of the target user is inquired in the risk knowledge graph, and the credit of the transfer object is further inquired in the risk knowledge graph. The description is given for illustrative purposes only and is not intended to be limiting.

As another example, the risk factor is the number of people who have overdue records among all objects with which the transfer transaction occurred with the target user. And inquiring users with overdue records in all the transfer objects of the target user in the risk knowledge graph, and recording the number of the users with the overdue records. The evaluation rule corresponding to the risk factor may include: within a preset time period, recording whether the first numerical value is exceeded or not; within a preset time period, recording whether the overdue record exceeds a second numerical value; and within a preset time period, recording whether the overdue record exceeds a third numerical value. Wherein the first value is less than the second value, and the second value is less than the third value. The description is given for illustrative purposes only and is not intended to be limiting.

Generally, if the target user frequently comes and goes with a person with low credit or overdue records, the credit of the target user is worth studying. Therefore, the credit degree and overdue records of the transfer object are used as risk factors, and the finally obtained risk index can accurately reflect the credit of the target user. And then carrying out risk early warning on the enterprise according to the risk index.

As another example, the risk factor is a risk merchant, which may include a merchant that may include a cash register service (e.g., cash register, credit card cash register, etc.) that provides cash register services to the user. And inquiring whether the target user has a transaction record with the risk merchant in the risk knowledge graph, wherein the transaction times are the same. The evaluation rule corresponding to the risk factor may include: whether the target user transacts with the risk merchant; if the target user has traded with the risk commercial tenants, whether the number of the risk commercial tenants traded with the target user exceeds a preset threshold value or not; whether the total transaction times with the risk merchants exceed the preset transaction times or not and the like. The description is given for illustrative purposes only and is not intended to be limiting.

Typically, if a target user frequently transacts with an at-risk merchant, the target user may frequently cash out. Therefore, the risk merchant is also used as a risk factor, so that the finally obtained risk index can accurately reflect the credit of the target user. And then carrying out risk early warning on the enterprise according to the risk index.

S1062: and calculating the corresponding score of each risk factor according to the weighted value and the evaluation rule respectively corresponding to each risk factor and the risk data.

Illustratively, the risk data corresponding to each risk factor is evaluated according to the evaluation rule, and a score corresponding to each risk data is obtained. And multiplying the score corresponding to each risk data by the weight value corresponding to the risk factor to obtain the score corresponding to each risk factor.

S1063: and determining the risk index according to the corresponding score of each risk factor.

Illustratively, when there is only one risk factor, the risk index may be represented by a score corresponding to that risk factor.

Illustratively, when the risk factors are multiple, the score corresponding to each risk factor is obtained, the sum of all the scores is calculated, and the sum of all the scores is used as the risk index of the target user.

For example, the risk factors include the number of transfers, overdue records, credit for the transfer object, the risk merchant, and the like. And if the scores corresponding to the risk factors are a, b, c and d, respectively, the risk index of the target user is recorded as the sum of a, b, c and d. The description is given for illustrative purposes only and is not intended to be limiting.

Optionally, in some possible implementations of the present application, the S1062 may include S10621 to S10623, which are specifically as follows:

s10621: and acquiring target risk data corresponding to the risk factors in the risk data for each risk factor.

For each risk factor, the target risk data refers to the data in the risk data that is associated with that risk factor. Illustratively, the risk data is first searched for target risk data associated with the risk factor.

For example, if the risk factor is a transaction mode, corresponding transfer data, consumption payment data and the like are searched in the risk data, and the data are used as target risk data corresponding to the risk factor. For another example, if the risk factor is a overdue record, corresponding overdue information is searched for in the risk data, and the information is used as the target risk data corresponding to the risk factor. The description is given for illustrative purposes only and is not intended to be limiting.

S10622: and evaluating the target risk data according to the evaluation rule corresponding to the risk factor to obtain an initial score corresponding to the target risk data.

Illustratively, the risk data corresponding to each risk factor is evaluated according to the evaluation rule, and an initial score corresponding to each target risk data is obtained.

For example, the risk factor is the number of transfer accounts, when the number of transfer accounts corresponding to the target user exceeds a first preset number of transfer accounts within a preset time period, the initial score corresponding to the target risk data is marked as A. And when the number of the transfer accounts corresponding to the target user exceeds a second preset number within the preset time period, recording the initial score corresponding to the target risk data as B. And when the number of the transfer accounts corresponding to the target user exceeds a third preset number within the preset time period, recording the initial score corresponding to the target risk data as C. The description is given for illustrative purposes only and is not intended to be limiting.

S10623: and calculating the product of the weight value corresponding to the risk factor and the initial score, and determining the score corresponding to the risk factor based on the product.

Illustratively, the initial score corresponding to each target risk data is multiplied by the weight value corresponding to the risk factor to obtain the score corresponding to each risk factor.

Continuing with the above example, the risk factor is the number of transfer accounts, when it is detected that the number of transfer accounts corresponding to the target user exceeds a first preset number within a preset time period, the initial score corresponding to the target risk data is recorded as a, the product of the initial score a and the weight value corresponding to the risk factor is calculated, and the product is recorded as the score corresponding to the risk factor.

And when the number of the transfer accounts corresponding to the target user exceeds a second preset number within a preset time period, recording the initial score corresponding to the target risk data as B, calculating the product of the initial score B and the weight value corresponding to the risk factor, and recording the product as the score corresponding to the risk factor.

And when the number of the transfer accounts corresponding to the target user exceeds a third preset number within a preset time period, recording the initial score corresponding to the target risk data as C, calculating the product of the initial score C and the weight value corresponding to the risk factor, and recording the product as the score corresponding to the risk factor.

In the above embodiment, an enterprise may set the weight values corresponding to different risk factors according to different services, so that the finally calculated score is more suitable for the service of the enterprise. For example, an enterprise may focus on the risk of which aspect, and the weight value of the risk factor for that aspect may be set higher.

Referring to fig. 3, fig. 3 is a flowchart illustrating a step S107 of a method for identifying risk according to another exemplary embodiment of the present application; optionally, in some possible implementations of the present application, the S107 may include S1071 to S1072, which are specifically as follows:

s1071: and determining the risk grade corresponding to the risk index.

Illustratively, different risk levels are set in advance according to different risk indexes.

Optionally, the lower the risk index, the higher the confidence justified for the target user; the higher the risk index, the lower the confidence that the target user is justified.

For example, when the risk index is within a first index range, one level of risk is assigned; when the risk index is in a second index range, corresponding to a second risk level; when the risk index is within the third index range, three levels of risk are assigned. Wherein the risk level of the first level of risk is lower than the second level of risk, and the risk level of the second level of risk is lower than the third level of risk.

For example, in the case of a loan transaction, the lower the risk index of a target user, the less likely it is that the target user will be overdue or unable to make a payment for making a loan or transacting a credit card. Conversely, the higher the risk index of the target user, the greater the likelihood that the target user will be overdue or unable to make a payment for the loan or credit card.

Optionally, in a possible implementation manner, the higher the risk index is, the lower the credit of the corresponding target user is; the lower the risk index, the higher the credit rating of the corresponding target user.

For example, when the risk index is within the first index range, three levels of risk are assigned; when the risk index is in a second index range, corresponding to a second risk level; when the risk index is within the third index range, the risk level is one level. And the risk level of the risk level III is lower than that of the risk level II, and the risk level of the risk level II is lower than that of the risk level I.

In this case, in the case of the loan transaction, the higher the risk index of the target user is, the higher the possibility that the target user may be overdue or unable to make a payment for making a loan or transacting a credit card for the target user is. Conversely, the lower the risk index of the target user, the less likely that the target user may be overdue or unable to make a payment for the loan or credit card transaction. The description is given for illustrative purposes only and is not intended to be limiting.

S1072: and carrying out risk early warning of different levels on the target enterprise according to the risk level.

Illustratively, different risk levels correspond to different levels of risk forewarning. For example, the higher the risk index, the lower the credit rating of the corresponding target user; the lower the risk index, the higher the credit rating of the corresponding target user in such a scenario.

When the risk level is first level, the risk early warning mode for the target enterprise can be information reminding. When the risk level is the second level, the risk early warning mode for the target enterprise can be voice reminding. When the risk level is three-level, the risk early warning mode for the target enterprise can be that the information is reminded and the voice is reminded at the same time. The information of the target user can be displayed in a striking color, so that the working personnel can attach importance to the target user, and economic loss of target enterprises is avoided. The description is given for illustrative purposes only and is not intended to be limiting.

Referring to FIG. 4, FIG. 4 is a flowchart illustrating a method for constructing a risk profile according to yet another exemplary embodiment of the present application; the embodiment of the present invention differs from the embodiment corresponding to fig. 1 in that after S207, S207 is further included, S201 to S207 in the present embodiment are completely the same as S101 to S107 in the embodiment corresponding to fig. 1, and reference is specifically made to the description related to S101 to S107 in the previous embodiment, which is not repeated herein. S208 is specifically as follows:

s208: and generating a risk log according to the various risk factors, the target user and the risk data.

Optionally, all risk data of the target user queried in the risk knowledge graph may be exported, risk data corresponding to each exported risk factor, each risk factor, and the target user are associated, and a risk log is generated.

The risk log is convenient for workers and target users to visually check various items of risk data of the target users, the risk indexes of the target users are searchable no matter whether the risk indexes are high or low, the current risk indexes are convenient to explain to the target users, and meanwhile the credibility of enterprises is increased.

Optionally, in some possible implementations of the present application, the method for identifying a risk provided by the present application may be applied in the medical field.

For example, when a medical loan is made, the method for identifying risks provided by the application can be used for predicting the risk index of a loan user and further helping the loan party to judge whether to pay the user.

For another example, when crowd funding is required for treating diseases, the risk index of the user who initiates crowd funding can be predicted through the risk identification method provided by the application. The prediction result can be disclosed to the public, and then the people are helped to judge whether to donate money for the user.

Referring to fig. 5, fig. 5 is a schematic view illustrating an apparatus for identifying risk according to an embodiment of the present disclosure. The device comprises units for performing the steps in the embodiments corresponding to fig. 1-4. Please refer to the related description of the embodiments corresponding to fig. 1 to 4.

For convenience of explanation, only the portions related to the present embodiment are shown. Referring to fig. 5, it includes:

an acquisition unit 310, configured to acquire target data for constructing a risk knowledge graph;

an extracting unit 320, configured to extract, from the acquired target data, a plurality of entities, attribute information of each entity, and relationship information between the entities, where the entities include users, enterprises, and businesses;

the constructing unit 330 is configured to construct an initial risk knowledge graph according to the extracted entities, the extracted attribute information of the entities, and the relationship information between the entities;

an obtaining unit 340, configured to obtain multiple risk factors of the user, import the multiple risk factors of the user into the initial risk knowledge graph, and generate the risk knowledge graph, where the multiple risk factors include a transaction manner, overdue records, credit of a transaction object, and a risk merchant;

a first determining unit 350, configured to determine risk data of a target user based on the risk knowledge-graph;

a second determining unit 360, configured to determine a risk index of the target user according to the risk data;

and the early warning unit 370 is configured to perform risk early warning on the target enterprise according to the risk index.

Optionally, the second determining unit 360 is specifically configured to:

Optionally, the second determining unit 360 is further configured to:

Optionally, the early warning unit 370 is specifically configured to:

determining a risk grade corresponding to the risk index;

Optionally, the apparatus further comprises:

the eliminating unit is used for eliminating the over-point data in the risk data, the over-point data comprises transaction data of the target user and white list merchants, and the white list merchants are merchants with more preset times of transactions with a plurality of users;

the second determining unit 360 is specifically configured to: and determining the risk index of the target user according to the risk data after the over point data is removed.

Optionally, the apparatus further comprises:

and the log generating unit is used for generating a risk log according to the risk factor, the target user and the risk data.

Optionally, the apparatus further comprises:

the conversion unit is used for converting the format of the target data into a JSON format;

the extracting unit 320 is specifically configured to: and extracting a plurality of entities, attribute information of each entity and relationship information among the entities from the target data in the JSON format.

Referring to fig. 6, fig. 6 is a schematic diagram of an apparatus for identifying risk according to another embodiment of the present application. As shown in fig. 6, the apparatus 4 for risk identification of this embodiment includes: a processor 40, a memory 41 and a computer program 42 stored in said memory 41 and executable on said processor 40. The processor 40, when executing the computer program 42, implements the steps in the various risk identification method embodiments described above, such as S101-S107 shown in fig. 1. Alternatively, the processor 40 implements the functions of the units in the above embodiments, such as the functions of the units 310 to 370 shown in fig. 5, when executing the computer program 42.

Illustratively, the computer program 42 may be divided into one or more units, which are stored in the memory 41 and executed by the processor 40 to accomplish the present application. The one or more units may be a series of computer instruction segments capable of performing specific functions, which are used to describe the execution of the computer program 42 in the device 4. For example, the computer program 42 may be divided into an acquisition unit, an extraction unit, a construction unit, an acquisition unit, a first determination unit, a second determination unit, and an early warning unit, each of which functions as described above.

The apparatus may include, but is not limited to, a processor 40, a memory 41. Those skilled in the art will appreciate that fig. 6 is merely an example of a device 4 and does not constitute a limitation of the device and may include more or fewer components than shown, or some components in combination, or different components, e.g., the device may also include input output devices, network access devices, buses, etc.

The processor 40 may be a Central Processing Unit (CPU), other general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The memory 41 may be an internal storage unit of the device, such as a hard disk or a memory of the device. The memory 41 may also be an external storage terminal of the device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like provided on the device. Further, the memory 41 may also include both an internal storage unit and an external storage terminal of the apparatus. The memory 41 is used for storing the computer instructions and other programs and data required by the terminal. The memory 41 may also be used to temporarily store data that has been output or is to be output.

The embodiments of the present application further provide a computer storage medium, where the computer storage medium may be non-volatile or volatile, and the computer storage medium stores a computer program, and when the computer program is executed by a processor, the computer program implements the steps in the above-mentioned method embodiments for identifying risks.

The present application also provides a computer program product which, when run on an apparatus, causes the apparatus to perform the steps in the various risk identifying method embodiments described above.

An embodiment of the present application further provides a chip or an integrated circuit, where the chip or the integrated circuit includes: a processor for calling and running the computer program from the memory so that the device on which the chip or integrated circuit is mounted performs the steps in the above-described method embodiments for identifying risks.

It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-mentioned division of the functional units and modules is illustrated, and in practical applications, the above-mentioned function distribution may be performed by different functional units and modules according to needs, that is, the internal structure of the apparatus is divided into different functional units or modules, so as to perform all or part of the functions described above. Each functional unit and module in the embodiments may be integrated in one processing unit, or each unit may exist alone physically, or two or more units are integrated in one unit, and the integrated unit may be implemented in a form of hardware, or in a form of software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working processes of the units and modules in the system may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and reference may be made to the related descriptions of other embodiments for parts that are not described or illustrated in a certain embodiment.

Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

The above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not cause the essential features of the corresponding technical solutions to depart from the spirit scope of the technical solutions of the embodiments of the present application, and are intended to be included within the scope of the present application.

Claims

1. A method of identifying risk, comprising:

acquiring target data for constructing a risk knowledge graph;

determining risk data of a target user based on the risk knowledge graph;

determining a risk index of the target user according to the risk data;

2. The method of claim 1, wherein said determining a risk index for the target user from the risk data comprises:

3. The method of claim 2, wherein said calculating a score corresponding to each of said risk factors according to the weight value and evaluation rule corresponding to each of said risk factors and said risk data comprises:

4. The method of claim 1, wherein the risk pre-warning the target enterprise according to the risk index comprises:

determining a risk grade corresponding to the risk index;

5. The method of claim 1, wherein prior to determining the risk index for the target user from the risk data, the method further comprises:

determining a risk index for the target user based on the risk data comprises:

6. The method of any one of claims 1 to 5, wherein after the risk pre-warning the target enterprise according to the risk index, the method further comprises:

7. The method of claim 1, wherein prior to extracting a plurality of entities, attribute information for each of the entities, and relationship information between the entities in the collected target data, the method further comprises:

converting the format of the target data into a JSON format;

8. An apparatus for identifying risk, comprising:

9. An apparatus for identifying a risk, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the method of any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1 to 7.