CN118657212A

CN118657212A - Trusted processing and knowledge graph processing method, device, equipment, medium and product

Info

Publication number: CN118657212A
Application number: CN202410770055.4A
Authority: CN
Inventors: 张彪; 吴楠; 黄德荣; 汤民帮; 曾欣; 程莹; 吕飞鹏
Original assignee: China Construction Bank Corp; CCB Finetech Co Ltd
Current assignee: China Construction Bank Corp; CCB Finetech Co Ltd
Priority date: 2024-06-14
Filing date: 2024-06-14
Publication date: 2024-09-17

Abstract

The application provides a trust processing and knowledge graph processing method, a trust processing and knowledge graph processing device, equipment, a medium and a product. Relates to the technical field of big data. The method comprises the following steps: acquiring data related to a plurality of objects and a target service in a banking system; according to the data related to the target business of a plurality of objects in the banking system and the information of the target business, constructing a knowledge graph of the target business by adopting a federal knowledge graph construction mode; reasoning the implicit relation between the entities in the knowledge graph of the target service; updating the implicit relation obtained by reasoning into the knowledge graph of the target service; constructing a knowledge graph of sub-businesses of each target business of each object based on the updated knowledge graph of the target business, and an entity weight set of the sub-businesses; the set of weights includes: the weight between entities in the knowledge graph corresponding to the sub-business. The method of the application improves the accuracy of the credit processing.

Description

Trusted processing and knowledge graph processing method, device, equipment, medium and product

Technical Field

The application relates to the technical field of big data, in particular to a trust processing and knowledge graph processing method, a device, equipment, a medium and a product.

Background

Because of the diversification of the objects in the bank credit service, the repayment credit and the capability of the objects in the credit service are difficult to guarantee. The accuracy of the credit giving processing method in the client credit giving, risk mining, post-credit early warning and monitoring links in the credit giving business is important for the bank to effectively manage risks, guarantee fund safety and improve the credit giving business efficiency.

In the prior art, a banking system generally extracts data related to a credit service from a text as user data, and evaluates the credit status of a user through a credit processing method based on the user data. The trust processing method generally comprises expert rules, expert scoring and scoring cards. The expert rule method relies on rules formulated by expert experience and knowledge to select part of data analysis from user data; the expert scoring method selects partial data analysis from the user data based on subjective judgment and experience of an expert; the scoring card method relies on feature data selected from user data to calculate a numeric score by building a mathematical model. Therefore, the existing credit granting processing method relies on expert experience and knowledge to select feature data and evaluate credit conditions of users, and accuracy of credit granting processing is reduced.

Disclosure of Invention

The application provides a credit processing and knowledge graph processing method, a device, equipment, a medium and a product, which are used for solving the problems that the accuracy of the credit processing is reduced by selecting characteristic data and evaluating the credit condition of a user according to expert experience and knowledge in the conventional credit processing method.

In a first aspect, the present application provides a knowledge graph processing method, including:

Acquiring data related to a plurality of objects and a target service in a banking system;

According to the data related to the target business of a plurality of objects in the banking system and the information of the target business, constructing a knowledge graph of the target business by adopting a federal knowledge graph construction mode; the information of the target service comprises: the domain to which the target business belongs, the activity content of the target business;

reasoning the implicit relation between the entities in the knowledge graph of the target service;

updating the implicit relation obtained by reasoning into the knowledge graph of the target service;

Constructing a knowledge graph of sub-businesses of each target business of each object based on the updated knowledge graph of the target business, and an entity weight set of the sub-businesses; the set of weights includes: the weight between entities in the knowledge graph corresponding to the sub-business.

In one possible design, according to data related to a target service by a plurality of objects in the banking system and information of the target service, a federal knowledge graph construction mode is adopted to construct a knowledge graph of the target service, including:

Constructing entity types of the knowledge graph of the target service according to the field of the target service;

Determining an entity relationship type according to the activity content of the target service;

Extracting an entity, an entity relation and an entity attribute related to the entity type from data related to a target service of the plurality of objects in a banking system according to the entity type and the relation type;

Resolving contradictions and ambiguities between the entity, the entity relationship and the entity attribute;

And constructing a knowledge graph of the target service by utilizing the entity, the entity relation and the entity attribute after contradiction and ambiguity elimination and adopting a federal knowledge graph construction mode.

In one possible design, constructing the entity weight set of the sub-service based on the updated knowledge-graph of the target service includes:

based on knowledge maps of a plurality of objects corresponding to the same sub-service, acquiring an entity weight set of the sub-service by using a graph algorithm.

In one possible design, after constructing the knowledge graph of the target service, the method includes:

detecting whether an object and/or data related to the object and a target service in the banking system change or not;

updating the knowledge graph of the target service in response to changes in the object and/or data related to the object and the target service in the banking system;

And updating the knowledge spectrum of the sub-business of each target business of each object and the entity weight set of the sub-business based on the updated knowledge spectrum of the target business.

In a second aspect, the present application provides a target service trust processing method, including:

Responding to a target sub-service credit service processing request of a target service of a target object, and acquiring a knowledge graph of the target object for the target sub-service from the knowledge graph of an object corresponding to the target service;

Acquiring an entity weight set of the target object from the entity weight set of the target sub-service based on the entity in the knowledge graph of the target object aiming at the target sub-service; the knowledge graph and the entity weight set of the target object aiming at the target sub-service are obtained based on the method according to any one of the first aspect;

acquiring a target sub-service credit service processing model from a credit service processing model library;

And acquiring a target sub-service credit giving processing result of the target object by using the target sub-service credit giving service processing model according to the knowledge graph and the entity weight set of the target object aiming at the target sub-service.

In a third aspect, the present application provides a knowledge-graph processing apparatus, including:

The data acquisition module is used for acquiring data related to a plurality of objects and target business in the banking system;

The first construction module is used for constructing a knowledge graph of the target service by adopting a federal knowledge graph construction mode according to data related to the target service and information of the target service of a plurality of objects in the banking system; the information of the target service comprises: the domain to which the target business belongs, the activity content of the target business;

The reasoning module is used for reasoning the implicit relation between the entities in the knowledge graph of the target service;

the updating module is used for updating the implicit relation obtained by reasoning into the knowledge graph of the target service;

The second construction module is used for constructing a knowledge graph of a sub-service of each target service of each object and an entity weight set of the sub-service based on the updated knowledge graph of the target service; the set of weights includes: the weight between entities in the knowledge graph corresponding to the sub-business.

In a fourth aspect, the present application provides a target service trust processing apparatus, including:

the knowledge graph acquisition module is used for responding to a target sub-service credit service processing request of a target service of a target object and acquiring a knowledge graph of the target object for the target sub-service from the knowledge graph of an object corresponding to the target service;

The acquisition weight set module is used for acquiring the entity weight set of the target object from the entity weight set of the target sub-service based on the entity in the knowledge graph of the target object for the target sub-service; the knowledge graph and the entity weight set of the target object aiming at the target sub-service are obtained based on the method according to any one of the first aspect;

The acquisition model module is used for acquiring a target credit service processing model from the credit service processing model library;

and the processing module is used for acquiring a target sub-service credit-giving processing result of the target object by utilizing the target sub-service credit-giving service processing model according to the knowledge graph and the entity weight set of the target object aiming at the target sub-service.

In a fifth aspect, an embodiment of the present application provides an electronic device, including: a processor, and a memory communicatively coupled to the processor;

the memory stores computer-executable instructions;

the processor executes computer-executable instructions stored by the memory to implement the method of any one of the first or second aspects.

In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium having stored therein computer-executable instructions which, when executed by a processor, implement a method as in any of the first or second aspects.

In a seventh aspect, an embodiment of the present application provides a computer program product comprising a computer program which, when executed by a processor, implements a method according to any of the first or second aspects.

The trust processing and knowledge graph processing method, the device, the equipment, the medium and the product provided by the application can acquire the execution sequence and the execution state of the transaction from the transaction state table when the transaction is processed by adding the transaction state table in the distributed system, so that the mode of processing the transaction is determined according to the execution sequence, and the consistency of the transaction is maintained. Based on the mode of the application, for the banking transaction system which does not realize the consistency of transactions in the prior period, the transaction consistency can be realized by only adding the processing logic of the transaction table in the banking transaction system without redesigning and splitting the transaction business and reconstructing the banking transaction system, the design difficulty for realizing the maintenance of the transaction consistency is reduced, and the realization efficiency is improved.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description, serve to explain the principles of the application.

FIG. 1 is a flowchart of a knowledge graph processing method according to an embodiment of the present application;

Fig. 2 is a flowchart of a method for constructing a knowledge graph of a target service according to an embodiment of the present application;

FIG. 3 is a flowchart of a method for constructing knowledge graphs of sub-services of each target service of each object according to an embodiment of the present application;

fig. 4 is a flowchart of a target service trust processing method provided in an embodiment of the present application;

Fig. 5 is a schematic structural diagram of a knowledge-graph processing apparatus according to an embodiment of the present application;

fig. 6 is a schematic structural diagram of a target service trust processing device according to an embodiment of the present application;

fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Specific embodiments of the present application have been shown by way of the above drawings and will be described in more detail below. The drawings and the written description are not intended to limit the scope of the inventive concepts in any way, but rather to illustrate the inventive concepts to those skilled in the art by reference to the specific embodiments.

Detailed Description

Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the application. Rather, they are merely examples of apparatus and methods consistent with aspects of the application as detailed in the accompanying claims.

In the technical scheme of the application, the related processes of collecting, storing, using, processing, transmitting, providing, disclosing and the like of the information such as financial data or user data are in accordance with the regulations of related laws and regulations, and the public welfare is not violated.

It should be noted that, in the embodiments of the present application, some existing solutions in the industry such as software, components, models, etc. may be mentioned, and they should be regarded as exemplary, only for illustrating the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or must not use the solution.

Because of the diversification of objects in the bank credit service, the objects in the credit service, namely individuals, are diversified in information forms such as financial information, professional information, family condition, expense and the like, and the economic condition of the individuals is complex, and the credit and capability of the individuals repayment in the credit service are difficult to guarantee. Therefore, the accuracy of the processing method of the client credit, risk mining and post-credit early warning monitoring links in the credit service is important for the bank to effectively manage the risk, ensure the fund safety and improve the credit service efficiency.

In the prior art, a banking system generally extracts data related to personal trust from a text as user data, and evaluates the credit status of a user through a trust processing method based on the user data. The trust processing method generally comprises expert rules, expert scoring and scoring cards. The expert rule method relies on rules formulated by expert experience and knowledge to select part of data analysis from user data; the expert scoring method selects partial data analysis from the user data based on subjective judgment and experience of an expert; the scoring card method relies on feature data selected from user data to calculate a numeric score by building a mathematical model. Therefore, the existing credit granting processing method relies on expert experience and knowledge to select feature data and evaluate credit conditions of users, and accuracy of credit granting processing is reduced.

Further analyzing the data used for trust in the prior art, the prior art generally selects part of the data from the user data as the data used for trust to carry out trust processing, wherein the user data comprises the explicit relationship of the user which is directly extracted from the information related to the user, and does not comprise the implicit relationship of the user. Therefore, the prior art only carries out the credit processing based on the dominant relationship of the users, and the accuracy of the credit processing is reduced.

In view of this, the present application provides a trust processing method, which obtains data for trust processing based on a constructed trust knowledge graph, and inputs the data for trust processing into a trust service processing model to obtain trust processing results of users. The knowledge graph processing method is provided for constructing the credit-giving knowledge graph, the knowledge graph is constructed based on the user data, so that the implicit relation in the user data is acquired through knowledge reasoning, the data for the credit-giving processing is acquired by integrating the explicit relation and the implicit relation of the user data, and the dimension of the data for the credit-giving processing is improved. Based on the trust knowledge graph processing and trust processing method provided by the application, trust processing is avoided depending on expert experience, and trust processing accuracy is improved.

Furthermore, the trust knowledge graph discovers the implicit relation of the user through knowledge reasoning, and can combine the explicit relation and the implicit relation of the user data to construct data for trust processing, so that the dimension of the data for trust processing is improved. The trust processing is carried out based on the data for trust processing, so that the accuracy of the trust processing is further improved.

That is, the present application includes two parts of content:

Part 1 (i.e. knowledge graph processing stage): and constructing a knowledge graph based on the user data, and mining the implicit relation of the user through knowledge reasoning to construct data for trust processing by combining the explicit relation and the implicit relation of the user data.

Part 2 (i.e. trusted processing stage): and acquiring data for credit processing based on the constructed knowledge graph, and inputting the data for credit processing into a sub-service credit processing model to acquire a credit processing result of the user.

It should be understood that the method provided by the embodiment of the application can be applied to any trusted application scenario. The 1 st part and the 2 nd part may be the same execution body or may be different execution bodies, which is not limited in the present application. For example, the execution subjects of the 1 st and the 2 nd are banking systems for trusted processing, or any electronic devices (e.g., servers) having processing functions. Or the execution main body of the 1 st part is a processing system for processing the knowledge graph, and the execution main body of the 2 nd part is a banking system for credit giving processing and the like.

It should be noted that the banking system or the processing system may be disposed on one electronic device in any environment (for example, disposed on one edge server in an edge environment separately), may be disposed in a cloud environment entirely, or may be disposed in a distributed manner in different environments.

For example, a banking system or processing system may be logically divided into multiple sections, each section having a different function. Portions of the banking system or processing system may be deployed in any two or three of the electronic device (on the user side), the edge environment, and the cloud environment, respectively. An edge environment is an environment that includes a collection of edge electronic devices that are closer to the electronic device, the edge electronic device comprising: edge servers, edge kiosks with computational power, etc. The various parts of the banking system or processing system deployed in different environments or devices cooperate to implement the functions of the banking system or processing system.

It should be understood that the present application does not carry out a restrictive division on what environments the bank system or the processing system is deployed in, and in practical application, the present application can carry out an adaptive deployment according to the computing power of the electronic device, the resource occupation situation of the edge environment and the cloud environment, or the specific application requirements.

In order to facilitate understanding, the technical scheme of the present application and how the technical scheme of the present application solves the above technical problems will be described in detail below by taking the execution body of the 1 st part as a processing system for processing a knowledge graph, and the execution body of the 2 nd part as a banking system for trusted processing. The following embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. Embodiments of the present application will be described below with reference to the accompanying drawings.

The following describes a knowledge graph processing method provided by the embodiment of the application. I.e. section 1 is explained first.

Fig. 1 is a flowchart of a knowledge graph processing method according to an embodiment of the present application. As shown in fig. 1, the method may for example comprise the steps of:

s101, acquiring data related to a target service of a plurality of objects in a banking system.

The plurality of objects refers to a plurality of individuals included in the user data. The user data may be obtained, for example, in a user database, which may be a database of user data included in a banking system.

The target transaction is a banking transaction associated with credit, which may be, for example, a personal credit, credit card application. Data related to the target business refers to personal financial information, professional information, social relationships, revenue status, asset status, credit records, etc. that may be used to evaluate the credit status of a user in the target business.

It should be appreciated that the data format included in the user database may be text, audio, video, etc., and the data associated with the target transaction for the plurality of objects obtained by the processing system from the banking system may be data in any data format. The trust processing is carried out based on the data with various data formats, so that a more-dimensional data base can be provided for the construction of the knowledge graph. Considering that data in a single text format can improve processing efficiency in subsequent data processing, the embodiment of the application is described by taking data acquired by a processing system from a user database of a banking system as a text type.

The processing system obtains data relating to the target business for a plurality of objects from a user database. One possible implementation is obtained by predefined keywords related to the target service. The keywords are keywords related to the target service, which are preset in the processing system by trusted processors, and the processing system acquires sentences or paragraphs containing the keywords from text data in a user database based on the preset keywords, and the sentences or paragraphs are used as data related to the target service for a plurality of objects.

One possible implementation way uses a text clustering model to obtain data of a plurality of objects related to a target service from text data. For example, the processing system may train a text clustering model using data related to the target business, the trained text clustering model may automatically cluster, and extract text segments related to the target business from the clustered results as data related to the target business for a plurality of objects.

And extracting data related to the target service of a plurality of objects by the partial data acquired by the processing system, so that the data volume required to be processed subsequently is reduced, and the efficiency of constructing the knowledge graph is improved.

S102, constructing a knowledge graph of the target service by adopting a federal knowledge graph construction mode according to data related to the target service and information of the target service of a plurality of objects in a banking system.

The information of the target service includes the domain to which the target service belongs and the active content of the target service. Taking the example that the target business is personal credit, the field to which the target business belongs may be credit, and the activity content of the target business may be transaction and income.

The federal knowledge graph is different from graphs of other knowledge types, has distributed and cross-organization properties, is a knowledge graph set maintained by a plurality of organizations, and is suitable for constructing the trust knowledge graph needing distributed management. The processing system may construct a knowledge-graph of the target business using the federal knowledge-graph. The knowledge graph of the target service refers to a knowledge graph constructed based on data related to the target service, and can be used as data of an analysis object in the trust processing.

Before the processing system builds the knowledge graph of the target service, the entity type and the relation type related to the information of the target service can be confirmed, and the entity, the entity relation and the entity attribute required for building the knowledge graph are determined from the acquired data related to the target service of a plurality of objects in the banking system according to the entity type and the relation type. Taking the example of constructing the federal knowledge-graph, the entity represents an individual or object in the data related to the target business, and may be, for example, "person", "place", "credit card", etc. Entity relationships are used to represent associations between entities, such as "owning," "living," "spouse," and the like. The entity attribute is characteristic information describing the entity, for example, "age", "gender", and the like.

A knowledge graph is a graphical structure that represents entities, entity relationships, and entity attributes as being composed of nodes and edges. The processing system can take the obtained entity as a node of the knowledge graph and take the obtained entity relationship as an edge connected between the nodes so as to construct the knowledge graph of the target service.

The processing system may use the entity relationship and the entity attribute as nodes after obtaining the entity, the entity relationship and the entity attribute, and edges of the nodes and the entity nodes of the entity attribute as the values of the entity attribute to add the entity attribute to the structure of the knowledge graph. Alternatively, the processing system may also store the entity attributes separately in a table so that the entity attributes can be quickly queried.

For example, the processing system may input the entity, entity relationship, and entity attribute required to determine the knowledge-graph to construct the target business into a tool for constructing the knowledge-graph, and the tool for constructing the knowledge-graph may construct the input entity, entity relationship, and entity attribute into the knowledge-graph of the target business.

S103, reasoning the implicit relation among the entities in the knowledge graph of the target service.

The implicit relationship is an implicit relationship between two entities obtained through reasoning, namely, the relationship obtained through reasoning on the basis of the explicit relationship. Inference is the operation of obtaining new relationships or attributes through knowledge reasoning by utilizing entities in the knowledge graph, entity relationships and entity attribute information.

The knowledge graph of the target service constructed by the processing system comprises data related to the target service, and knowledge reasoning can be performed based on the knowledge graph of the target service so as to mine the implicit relation between entities in the knowledge graph of the target service. During the trust processing, the implicit relationship and the explicit relationship analysis of the user are combined, so that the association between data can be included in a multidimensional manner, abnormal conditions and potential risks of the data can be found more comprehensively, and the accuracy of the trust analysis is improved.

In one possible implementation manner, the knowledge reasoning may be based on a rule reasoning method, and the rule may be a logic rule preset by a trust processor. Under the implementation mode, the trust processor pre-sets the logical expression of the reasoning, and further utilizes the reasoning engine to carry out logical reasoning according to the reasoning rule of the logical expression so as to deduce the hidden relation in the knowledge graph of the target service.

One possible implementation manner can be to perform knowledge reasoning based on a graph algorithm, input the content of the knowledge graph of the target service into the graph algorithm, analyze paths among nodes in the knowledge graph of the target service through path searching, graph traversing and the like, and output a hidden relationship obtained by reasoning.

S104, updating the implicit relation obtained by reasoning into the knowledge graph of the target service.

And obtaining the implicit relation between the entities through knowledge reasoning of the knowledge graph of the target service. The processing system can update the obtained implicit relation into the knowledge graph of the target service, namely, new relation or attribute obtained by reasoning is added into the knowledge graph of the target service in a pattern structure mode. The implicit relation obtained by reasoning is a logic expression, and the implicit relation needs to be converted into nodes and edges in a graph structure of the knowledge graph, and the method for constructing the knowledge graph of the target service can be specifically referred to S102.

S105, constructing a knowledge graph of sub-business of each target business of each object and a business entity weight set of the sub-business based on the updated knowledge graph of the target business.

The weight set comprises weights among entities in the knowledge graph of the sub-business corresponding to the target business.

The knowledge graph of the sub-business of the target business is a sub-knowledge graph extracted from the knowledge graph of the target business according to the sub-business to be analyzed. The sub-business to be analyzed is any link of the target business, taking the target business as personal credit as an example, and the sub-business can be client credit giving, risk mining, post-credit early warning monitoring and the like.

The knowledge graph of the sub-business of each target business of each object is a sub-knowledge graph extracted for each object and each sub-business to be analyzed. The object is an entity with a person as a human in the knowledge graph of the target business. For example, if the sub-services to be analyzed include client trust and risk mining, the knowledge graph of the client trust sub-service of the object and the knowledge graph of the risk mining sub-service of the object are extracted for each object respectively.

The sub-services of different target services, the information emphasis to be analyzed is different. Taking the example that the target business is personal credit and the sub-business to be analyzed is client credit, the client credit refers to the link of credit decision making to the client before the bank provides the credit product, so the information focused by the sub-business includes financial condition, credit record and the like. Taking post-loan early warning and monitoring as an example, the sub-business to be analyzed refers to a link of risk monitoring after a customer obtains a loan. Thus, information that is of particular interest to this sub-business includes repayment status, funds flow, etc.

According to the information focused by the sub-business, the related entities and entity relations are different. Taking a financial situation as an example, the entities and entity relationships related thereto may be "real estate", "possession", etc.

And constructing a knowledge graph of sub-business of each target business of each object. The processing system may use the entity of each object as a central node, and use the related entity and entity relationship of the information focused by the sub-services of different target services as extraction parameters, to extract the knowledge graphs of the sub-services of the corresponding target services from the knowledge graphs of the target services.

For example, a graph database or graph analysis tool may be used, with which the center node and the extraction parameters are input, and the knowledge graph of the sub-service of the target service related to the center node, that is, the knowledge graph of the sub-service of the target service of the object, is extracted. Optionally, the processing system may further input an extraction layer number, and the graph database or the graph analysis tool may further extract only nodes and relationships within a set extraction layer number range as a knowledge graph of a sub-service of the target service based on the extraction layer number.

The processing system takes the sub-business of the target business as the basis, and obtains the weight set of the relation between the entities in the knowledge graph of the sub-business of the target business corresponding to the sub-business based on the knowledge graph of the sub-business of the target business of the same sub-business of each object. The weight of the relationship between the entities characterizes the importance degree of the relationship between the entities for analyzing the sub-business of the target business.

Illustratively, the processing system may calculate the relationship weights between knowledge-graph entities of the sub-business based on a graph algorithm. For example, the processing system uses a graph algorithm capable of calculating the relation weight between the knowledge graph entities of the sub-services, takes the knowledge graph of the sub-services of each object of the same sub-service as input data, and outputs the relation weight between the entities in the knowledge graph of the sub-service corresponding to the sub-service according to the graph structure of the input knowledge graph of the sub-service. Or the processing system may also select an appropriate learning model, such as a support vector machine or a convolutional neural network, to pre-learn weights of the relationships to obtain a set of relationship weights between entities of the knowledge graph of the sub-business.

The acquired knowledge graph of the sub-business of the target business of the object comprises data of the object which can be used for analyzing the sub-business, wherein the entity and the entity relationship comprise an explicit relationship extracted from user data and an implicit relationship acquired through knowledge reasoning. The obtained entity weight set of the sub-service comprises weights which can be used for analyzing the relation between the related entities of the sub-service. Therefore, the acquired knowledge spectrum of the sub-business of the target business of the object and the sub-business entity weight set can be combined to analyze the personal credit condition of the object in the sub-business of the current target business so as to carry out credit giving processing.

According to the knowledge graph processing method, the knowledge graph of the target service is built based on the user data, the implicit relation in the user data is obtained through knowledge reasoning, the data used for the credit processing is obtained by integrating the explicit relation and the implicit relation of the user data, the dimension of the data used for the credit processing is improved, and compared with the method relying on expert experience in the prior art, the knowledge graph of the built target service can provide more comprehensive user data to carry out the credit processing, and the accuracy of the credit processing is improved.

The following illustrates how to construct a specific implementation of a knowledge graph of a target service by adopting a federal knowledge graph construction mode according to data related to the target service and information of the target service of a plurality of objects in a processing system.

Fig. 2 is a flowchart of a method for constructing a knowledge graph of a target service according to an embodiment of the present application. Alternatively, the knowledge graph of the target business may be constructed based on patterns (schemas). Schema provides a structured way to organize and describe knowledge, describing the process of designing and constructing a knowledge graph. The format of the knowledge graph of the constructed target service based on the Schema design is more standard, so that the query is easier. As shown in fig. 2, the method may for example comprise the steps of:

S201, constructing entity types of the knowledge graph of the target service according to the domain to which the target service belongs.

The entity type is a classification of the entity. For example, entity types may include "persona," "city," "credit card," and the like.

Taking the target business as personal credit and the knowledge graph to be constructed as credit knowledge graph as an example, the field to which the target business belongs can be credit. The processing system constructs entity types of credit knowledge graph according to the data of a plurality of objects related to the credit obtained from the banking system.

In one possible implementation manner, the entity type of the knowledge graph for constructing the target service may be an entity type preset in the processing system by the credit trust processor, and the preset entity type may be determined by the credit trust processor according to knowledge and experience of the domain to which the target service belongs.

In one possible implementation manner, the entity type for constructing the knowledge graph of the target service may be obtained based on an existing extraction method capable of extracting the entity type. The extraction method can be based on exploring and analyzing the related data of the field of the target business, analyzing the entity with the association, and classifying and extracting the entity type.

S202, determining the entity relation type according to the activity content of the target service.

Entity relationship types are relationships that exist between entities and may include, for example, "child," "spouse," "own," "transfer," and the like.

Taking the target business as personal credit and the knowledge graph to be constructed as credit trust knowledge graph as an example, the activity content of the target business can be transaction, income and the like. And the processing system acquires entity relation types constructing credit authorization knowledge graphs from the data related to the credit of a plurality of objects acquired from the banking system according to the activity content of target businesses such as transactions, transfer accounts and the like. The specific manner of acquiring the entity relationship type may refer to S201 for determining the method of the entity type.

S203, extracting the entity, the entity relation and the entity attribute related to the entity type from the data related to the target business of the plurality of objects in the banking system according to the entity type and the relation type.

The processing system can extract the entity, the entity relation and the entity attribute related to the entity type from the data related to the target service in the banking system by a plurality of objects according to the acquired entity type and relation type based on a text processing method, so as to be used for constructing a knowledge graph of the target service.

In one possible implementation, the processing system may identify, based on a named entity identification technique in the text processing method, an entity related to the entity type from the data related to the target service, and then extract, based on a relationship extraction technique such as dependency syntax analysis or semantic role labeling, an entity relationship between the entities from the data related to the target service, and may acquire, using the information extraction technique, an entity attribute describing entity information.

In one possible implementation, a processing system may extract entities, entity relationships, and entity attributes associated with entity types using prior knowledge extraction methods. For example, the processing system sets the entity type and the relation type to be extracted for the knowledge extraction method, and the knowledge extraction method can automatically extract the entity and the entity relation matched with the entity type and the relation type from the data by using machine learning, natural language processing and the like, and further automatically extract the entity attribute from the data based on the extracted entity.

S204, resolving contradictions and ambiguity for the entity, the entity relation and the entity attribute.

The extracted entity, entity relation and entity attribute related to the entity type may have the problem that the description is contradictory or ambiguous, and the processing system can process the entity, entity relation and entity attribute to eliminate the contradiction and ambiguity, so that the data for constructing the knowledge graph is more accurate.

For example, in one possible scenario, it is extracted that there are synonyms for an entity, an entity relationship, and an entity attribute, e.g., a "spouse," "couple," and "couple" in an entity relationship are all synonyms, and because the inconsistent expression language may cause duplication or ambiguity, the processing system may replace the synonyms with specified term expressions based on a synonym dictionary or a natural language processing method in the prior art to disambiguate.

In another possible case, the extracted entity, entity relationship and entity attribute have contradictory entity relationship, for example, the obviously contradictory entity relationship "child" and "grandchild" are extracted between the two entities, in which case, the processing system may correct the wrong entity relationship based on the natural language processing method in the prior art, or may also eliminate the contradiction by adopting a manner of deleting the contradictory entity relationship.

It should be understood that the above is merely illustrative of two methods for resolving contradictions and ambiguities between entities, entity relationships, and entity attributes, and that other methods may be employed to resolve contradictions and ambiguities between entities, entity relationships, and entity attributes, which the present application is not limited to.

S205, constructing a knowledge graph of the target service by utilizing the entity, the entity relation and the entity attribute after contradiction and ambiguity elimination and adopting a federal knowledge graph construction mode.

The knowledge graph is a structured knowledge representation, and represents entities, entity relationships and entity attributes as a graph structure consisting of nodes and edges. Federal knowledge-graph may include relationships of multiple entity types. For example, the knowledge graph of the target business constructed by using the federal knowledge graph can include relationships among various entity types such as individuals, places, individuals, credit cards and the like, wherein the relationship among each entity type can be regarded as a layer of sub-knowledge graph, that is, the federal knowledge graph comprises knowledge graphs of multiple layers of sub-knowledge graphs, and the included user data is more comprehensive. Further, the knowledge graph of the target business realized based on the federal knowledge graph is beneficial to distributed management and maintenance in each bank. Therefore, in the embodiment of the application, a federal knowledge graph construction mode is taken as an example to construct a knowledge graph of a target service, and the knowledge graphs mentioned in the following are federal knowledge graphs.

The processing system may use the obtained entity as a node of the knowledge graph, use the obtained entity relationship as an edge connected between the nodes of the entity, use the entity attribute value as a node, and use the entity attribute as an edge of the node of the attribute value and the node of the entity to construct the knowledge graph of the target service.

The processing system may input the entity, entity relationship, and entity attribute into a tool for building a knowledge-graph to obtain a knowledge-graph of the target business. For example, if the tool for constructing the knowledge graph is Neo4j, the processing system constructs and stores the federal knowledge graph in Neo4j by using the entity, the entity relationship and the entity attribute based on the graph structure, and uses the visualization tool provided by Neo4j as the knowledge graph of the target service, so as to visually reflect the structure and the association of the knowledge graph.

S206, reasoning the implicit relation among the entities in the knowledge graph of the target service.

The processing system may perform knowledge reasoning using a network analysis algorithm, for example, the processing system may transmit the knowledge graph to the network analysis algorithm in a format of graph data, and the network analysis algorithm obtains an implicit relationship between entities in the knowledge graph through knowledge reasoning. The network analysis algorithm may be, for example, graph loop analysis, multi-layer topology analysis, abnormal entity N degree analysis, or the like.

For example, for the explicit relationship "A is the spouse of B" and "C is the child of A" described in the text, A, B, C is the entity, "spouse", "child" is the relationship. When the knowledge graph is constructed, the entity A, B, C is respectively used as a node A, a node B and a node C in the knowledge graph, the node A is connected with the node B by taking the "spouse" as an edge, and the node A is connected with the node C by taking the "child" as an edge. After knowledge reasoning based on the knowledge graph, the implicit relation 'C is a child of B' which is not directly expressed in the text can be mined.

Furthermore, contradiction and ambiguity can be eliminated on the implicit relation obtained by reasoning so as to ensure the accuracy of the reasoning result. The processing system can process the implicit relation obtained by reasoning and the explicit relation in the knowledge graph to eliminate contradiction and ambiguity.

For example, the relationship between two inferred entities is "ancestor" and the explicit relationship of "child" exists between two inferred entities in the knowledge graph, that is, the implicit relationship obtained by inference contradicts the explicit relationship in the knowledge graph, so the contradiction can be eliminated by referring to the method for eliminating contradiction in S204, for example, the method for processing natural language in the prior art is adopted to eliminate contradiction.

Or the expression of the entity relationship in the implicit relationship may be replaced with a prescribed term expression with reference to the method of step S204 to eliminate the ambiguity problem caused by the inconsistent terms of the expression.

S207, updating the implicit relation obtained by reasoning into the knowledge graph of the target service.

Illustratively, the processing system may update the implicit relationship to the knowledge-graph of the target business using Neo4 j.

Based on the steps, a trust knowledge graph is constructed according to the data related to the target service of a plurality of objects in the processing system. The constructed knowledge graph comprises an explicit relation based on data extraction related to the target service and an implicit relation based on credit knowledge graph analysis, so that more-dimensional data can be provided for credit processing, and a processing result is more accurate.

The following illustrates how to construct a specific implementation of a knowledge graph of a sub-service of each target service of each object based on the knowledge graph of the target service in the knowledge graph processing method of the target service.

Fig. 3 is a flowchart of a method for constructing a knowledge graph of a sub-service of each target service of each object according to an embodiment of the present application. As shown in fig. 3, the method may for example comprise the steps of:

S301, acquiring entity types and entity relation types required by sub-businesses of each target business of a banking system.

The sub-services of different target services, the information emphasis to be analyzed is different. According to the information focused by the sub-business of the target business, the related entity types and entity relation types are different. Taking the information focused by the sub-business of the target business as the financial condition as an example, the entity type and the entity relation type related to the information can be 'real estate', 'possession', 'credit card', and the like.

The entity types and entity relationship types required by the sub-business of each target business of the banking system can be preset in the processing system. The preset entity type and entity relation type may be the entity type and entity relation type determined by the trust processor according to knowledge and experience to obtain the information emphasis point of the sub-service of each target service and the related data of the sub-service of each target service.

Or the processing system may be obtained based on existing extraction methods that can extract entity types and entity relationship types. The extraction method can be used for respectively searching based on the data of the sub-businesses of each target business, analyzing the related entity and entity relationship, classifying and extracting the entity type and entity relationship type of the sub-business of each target business, and presetting in a processing system.

The processing system obtains the entity type and the entity relation type of the sub-business corresponding to the target business from the entity type and the entity relation type of the sub-business of each target business preset in the processing system according to the sub-business of the target business.

S302, extracting the knowledge graph of the sub-business of each target business of each object from the updated knowledge graph of the target business based on the entity type and the entity relation type required by the sub-business of each target business of the bank system.

The target business is banking business related to credit, and the analysis object of the credit is usually a person, so that the processing system can extract the knowledge graph of the sub-business of each object by taking each object as a central node for analyzing the credit data of the person.

Because the sub-services of different target services have different emphasis on information to be analyzed, when each object is analyzed based on different sub-services, the knowledge graph of the sub-services can comprise different entity types and entity relationships, so that the correlation between the analyzed data and the sub-services is higher.

Therefore, the processing system can extract the knowledge graph of each sub-business of each object from the updated knowledge graph of each target business based on the entity type and the entity relationship type required by the sub-business of each target business of the banking system. For example, if the sub-services of the target service include client credit, risk mining, and post-loan early warning monitoring, the knowledge graph of the sub-service extracted for each object includes the knowledge graph of the client credit sub-service, the knowledge graph of the risk mining sub-service, and the knowledge graph of the post-loan early warning monitoring sub-service.

The processing system can respectively extract the knowledge graphs of the corresponding sub-businesses from the knowledge graphs of the target businesses by taking the entity of each object as a central node according to the entity type and the entity relation type required by each sub-business of the processing system. For example, the processing system may use a graph database or a graph analysis tool, take the object as a central node, take the entity type and the entity relationship type required by each sub-service as extraction parameters, input the extraction parameters into the graph database or the graph analysis tool, and extract a sub-knowledge graph related to the central node, that is, a knowledge graph of the sub-service of the object.

S303, acquiring a sub-business entity weight set by using a graph algorithm based on knowledge maps of a plurality of objects corresponding to the same sub-business.

The sub-business entity weight set comprises a weight set of the relation between the entities in the knowledge graph corresponding to the sub-business. Different entity weight sets are arranged corresponding to different sub-services, and the weights of the relationships among the entities included in the different sub-service entity weight sets are different, so that the larger the weights are, the larger the influence of the relationships among the corresponding entities on the sub-services is represented. For example, for an entity weight set in which a sub-service is a sub-service trusted by a client, the relationship between entities that characterizes financial conditions such as "possession", "transfer" or the like has a greater influence on client trust, and the weight of such entity relationship in the entity weight set will be greater.

The processing system may obtain the sub-business entity weight set using a graph algorithm based on knowledge maps of sub-business sub-categories of a plurality of objects of the same sub-business. For example, the processing system may input knowledge graphs of sub-services of multiple objects of the same sub-service into a graph algorithm, where the graph algorithm uses factors such as similarity between nodes, path length, node importance, and the like, and integrates node features, edge features, and global features to accurately obtain a weight value of each relationship as the sub-service entity weight set.

Optionally, when the sub-business entity weight set is acquired by using the graph algorithm, the knowledge graph of the sub-business of the plurality of objects may be the knowledge graph of the sub-business of all the objects of the same sub-business, or the knowledge graph of the sub-business of a part of the objects may be selected based on the attributes of the objects. Under the realization mode of selecting the knowledge graph of the sub-business of part of the objects, a more accurate customer group of the sub-business can be selected so as to improve the accuracy of the weights. For example, when the entity weight set of the knowledge graph of the client credit sub-service is performed, if the "age" attribute of the object is smaller than "18", part of the object is deleted based on the "age" attribute, so that the analyzed knowledge graph data of the client credit sub-service is more accurate, and the weight accuracy is further improved.

Based on the constructed knowledge graph of the target service, the processing system can also update nodes and edges in the knowledge graph of the target service. The processing system can detect whether the object and/or the data related to the object and the target service in the banking system change or not, and update the knowledge graph of the target service in response to the change of the object and/or the data related to the object and the target service in the banking system.

The processing system updates the knowledge graph of the target service, which may be to detect the change of the object and/or the data related to the target service in the bank system in response to the operation of the trusted processor, or may be to detect the change of the object and/or the data related to the target service in the bank system based on the period set in the processing system, which is not limited in this embodiment of the present application.

The processing system may obtain, according to a preset period in the processing system, change data of the object and/or data related to the object and the target service in the period time, and extract the entity and the entity relationship in the changed data, and add, modify, and delete corresponding nodes and edges in the knowledge graph of the target service in a graphic structure manner by using a tool capable of updating the knowledge graph.

According to the knowledge graph processing method, the knowledge graph of the target service is constructed based on the user data, the implicit relation of the user is mined through the characteristic that the knowledge graph can be used for knowledge reasoning, and the dominant relation and the implicit relation of the user data are combined, so that the dimension of the data for trust processing is improved. Further, the processing system performs the trust processing based on the provided data, so that the accuracy of the trust processing is improved.

The above-mentioned knowledge graph processing method according to the embodiment of the present application is based on the knowledge graph processing method, and each object may obtain a knowledge graph of a sub-service of each target service. The knowledge graph of each sub-service of each object can be stored in a trusted service processing database. The trusted service processing database can be a relational database, and the relational database can be stored in the database in a table form, so that the data can be conveniently queried.

For example, an object table, a node table, an edge table, and an attribute table may be created in the relational database to store knowledge maps of each sub-service of each object. The object table is used for storing a node table, an edge table and an attribute table corresponding to each sub-service knowledge graph of each object. The node table is used for storing node information in the knowledge graph, and each node corresponds to one row of records in the table. The edge table stores information of edges between the nodes, each edge corresponds to a row record in the table, and each row record contains nodes at two ends of edge connection and information of the edges. The attribute table is used for storing the attributes of each node, and the attribute of each node corresponds to one row of records in the table.

The following describes a target service trust processing method provided by the embodiment of the application. I.e. part 2. Fig. 4 is a flowchart of a target service trust processing method provided in an embodiment of the present application. The bank system for performing the target service trust processing is taken as an execution main body, as shown in fig. 4, and the target service trust processing method specifically comprises the following steps:

S401, responding to a target sub-service credit service processing request of a target service of a target object, and acquiring a knowledge graph of the target object for the target sub-service from the knowledge graph of the object corresponding to the target service.

The target object is an object to be analyzed by target service trust. The target sub-service of the target service is a sub-service to be processed, taking the target service as personal credit as an example, and the target sub-service can be client credit, for example. The target object stores the knowledge graph of the target sub-business in the trusted business processing database.

When processing the target sub-service credit service of the target service, the bank system queries and acquires the knowledge graph of the target sub-service of the target object from the knowledge graph of the sub-service stored in the credit service processing database according to the target object and the target sub-service in response to the target sub-service processing request for the target object for analyzing the target object.

S402, acquiring an entity weight set of the target object from the entity weight set of the target sub-service based on the entity in the knowledge graph of the target object for the target sub-service.

The knowledge graph processing method of the embodiment of the application obtains the weight set of each sub-business entity. The target sub-business entity weight set can be inquired from the stored sub-business entity weight sets according to the target sub-business.

The banking system can acquire the included entity and entity relation from the knowledge graph of the target sub-business of the target object, and further acquire the corresponding weight from the target sub-business entity weight set according to the acquired entity relation, and the corresponding weight is taken as the entity weight set of the target object.

S403, acquiring a target sub-service credit service processing model from a credit service processing model library.

The trusted service processing model library is a model library for storing the trusted service processing models of all sub-services in the banking system. The sub-business credit giving business processing models in the credit giving business processing model library are a plurality of sub-business credit giving business processing models which are obtained through training in advance according to the data of each sub-business.

For example, for the client trust sub-service, the entity relationship and the entity attribute included in the knowledge maps of the plurality of client trust sub-services may be pre-based, and the weight set of the relationship between the entities, and the client trust line may be used as training data to train the sub-service trust service processing model, and the trained sub-service trust service processing model may obtain the trust line of the target object based on the input relevant data of the knowledge maps of the client trust sub-service of the target object. The sub-service trust service processing model may be, for example, a limit gradient lifting (Extreme Gradient Boosting, XGBoost) model, a lightweight gradient lifting Machine (LIGHT GRADIENT Boosting Machine, lightGBM) model, or the like.

S404, acquiring a target sub-service credit giving processing result of the target object by utilizing the target sub-service credit giving service processing model according to the knowledge graph and the entity weight set of the target object aiming at the target sub-service.

The banking system can input the acquired target sub-business credit service processing model according to the entity, entity relation and entity attribute in the knowledge graph of the target sub-business of the target object and the acquired entity weight set of the target object as input data, output the credit service processing result of the target object, and the credit service processing result of the target object is determined according to the target credit service. Taking the target sub-service of the target service as the client credit giving example, when the client credit giving sub-service is processed, the banking system obtains the credit giving limit of the target object by using the client credit giving sub-service credit giving service processing model; for the post-credit early warning monitoring sub-business, the bank system can acquire the target sub-business credit processing result of the target object by utilizing the post-credit early warning monitoring sub-business credit processing model, and the target sub-business credit processing result can be the early warning level of the target object. The early warning level characterizes the possible repayment risk degree of the target object, for example, the early warning level can be an A level representing that the target object has no repayment risk, a C level representing that the target object has repayment risk, and the like.

According to the knowledge graph processing method, the knowledge graph of the target service is constructed based on the user data, the implicit relation of the user is mined through the characteristic that the knowledge graph can be used for knowledge reasoning, and the dominant relation and the implicit relation of the user data are combined, so that the dimension of the data for trust processing is improved. Furthermore, the target service credit processing method carries out credit processing based on the data provided by the knowledge graph, and because the data referred by the credit processing has higher dimensionality, the credit can comprehensively analyze the information of the object, wherein the information comprises the information of other objects related to the object, thereby improving the accuracy of the credit processing.

The foregoing is a description of an embodiment of the method of the present application, and the following describes an apparatus provided in the embodiment of the present application.

Fig. 5 is a schematic structural diagram of a knowledge-graph processing apparatus according to an embodiment of the present application. As shown in fig. 5, the apparatus may include, for example: an acquisition data module 501, a first building module 502, an inference module 503, an update module 504, a second building module 505. Optionally, a detection update module 506 may also be included.

An acquiring data module 501, configured to acquire data related to a target service from a plurality of objects in a banking system;

The first construction module 502 is configured to construct a knowledge graph of the target service according to data related to the target service and information of the target service by using a federal knowledge graph construction manner; the information of the target service includes: the domain to which the target business belongs, the activity content of the target business;

An inference module 503, configured to infer an implicit relationship between entities in a knowledge graph of a target service;

The updating module 504 is configured to update the implicit relationship obtained by reasoning to a knowledge graph of the target service;

A second construction module 505, configured to construct a knowledge graph of sub-services of each target service of each object and an entity weight set of the sub-services based on the updated knowledge graph of the target service; the weight set includes: the weight between entities in the knowledge graph corresponding to the sub-business.

In one possible implementation manner, the first construction module 502 is further configured to construct an entity type of the knowledge graph of the target service according to the domain to which the target service belongs; determining an entity relationship type according to the activity content of the target service; extracting an entity, an entity relation and an entity attribute related to the entity type from data related to a target service of a plurality of objects in a banking system according to the entity type and the relation type; resolving contradictions and ambiguity for entities, entity relationships and entity attributes; and constructing a knowledge graph of the target service by utilizing the entity, the entity relation and the entity attribute after contradiction and ambiguity elimination and adopting a federal knowledge graph construction mode.

A possible implementation manner, the second construction module 505 is specifically configured to obtain, using a graph algorithm, a set of entity weights of the sub-service based on knowledge maps of a plurality of objects corresponding to the same sub-service.

A possible implementation manner, the detection update module 506 is specifically configured to detect whether an object in the banking system and/or data related to the object and the target service change; updating the knowledge graph of the target service in response to changes in the object and/or data related to the object and the target service in the banking system; based on the updated knowledge graph of the target service, the knowledge graph of the sub-service of each target service of each object and the entity weight set of the sub-service are updated.

Fig. 6 is a schematic structural diagram of a target service trust processing device according to an embodiment of the present application. As shown in fig. 6, the apparatus may include, for example: the system comprises a knowledge graph acquisition module 601, a weight set acquisition module 602, a model acquisition module 603 and a processing module 604.

The knowledge graph acquisition module 601 is configured to acquire a knowledge graph of a target object for a target sub-service from a knowledge graph of an object corresponding to the target service in response to a target sub-service credit service processing request of the target service for the target object;

The acquiring weight set module 602 is configured to acquire an entity weight set of the target object from the entity weight sets of the target sub-service based on the entity in the knowledge graph of the target object for the target sub-service; the knowledge graph and entity weight set of the target object aiming at the target sub-business are obtained based on the knowledge graph processing method;

The obtaining model module 603 is configured to obtain a target trusted service processing model from a trusted service processing model library;

and the processing module 604 is configured to obtain a target sub-service credit-giving processing result of the target object by using the target sub-service credit-giving service processing model according to the knowledge graph and the entity weight set of the target object for the target sub-service.

It will be appreciated that the device embodiments described above are merely illustrative and that the device of the application may be implemented in other ways. For example, the division of the units/modules in the above embodiments is merely a logic function division, and there may be another division manner in actual implementation. For example, multiple units, modules, or components may be combined, or may be integrated into another system, or some features may be omitted or not performed.

Fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 7, the electronic device may include: at least one processor 701, a memory 702.

A memory 702 for storing a program. In particular, the program may include program code including computer-operating instructions.

The memory 702 may comprise high-speed RAM memory or may further comprise non-volatile memory (non-volatile memory), such as at least one disk memory.

The processor 701 is configured to execute computer-executable instructions stored in the memory 702 to implement the method of the foregoing method embodiment. The processor 701 may be a central processing unit (Central Processing Unit, CPU), or an Application SPECIFIC INTEGRATED Circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present application.

Optionally, the electronic device may also include a communication interface 703. In a specific implementation, if the communication interface 703, the memory 702, and the processor 701 are implemented independently, the communication interface 703, the memory 702, and the processor 701 may be connected to each other and perform communication with each other through buses.

Alternatively, in a specific implementation, if the communication interface 703, the memory 702, and the processor 701 are implemented on a single chip, the communication interface 703, the memory 702, and the processor 701 may complete communication through internal interfaces.

The present application also provides a computer-readable storage medium, which may include: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, and specifically, the computer readable storage medium stores program instructions for implementing the operations of the above method embodiments.

The present application also provides a program product comprising execution instructions stored in a readable storage medium. The at least one processor of the electronic device may read the execution instructions from the readable storage medium, the execution instructions being executed by the at least one processor to cause the electronic device to perform the actions of the method embodiments described above.

In addition, each functional unit/module in each embodiment of the present application may be integrated into one unit/module, or each unit/module may exist alone physically, or two or more units/modules may be integrated together, unless otherwise specified. The integrated units/modules described above may be implemented either in hardware or in software program modules.

The integrated units/modules, if implemented in hardware, may be digital circuits, analog circuits, etc. Physical implementations of hardware structures include, but are not limited to, transistors, memristors, and the like. The processor may be any suitable hardware processor, such as CPU, GPU, FPGA, DSP and an ASIC, etc., unless otherwise specified. Unless otherwise indicated, the storage elements may be any suitable magnetic or magneto-optical storage medium, such as resistive Random Access Memory RRAM (Resistive Random Access Memory), dynamic Random Access Memory DRAM (Dynamic Random Access Memory), static Random Access Memory SRAM (Static Random-Access Memory), enhanced dynamic Random Access Memory EDRAM (ENHANCED DYNAMIC Random Access Memory), high-Bandwidth Memory HBM (High-Bandwidth Memory), hybrid storage cube HMC (Hybrid Memory Cube), etc.

The integrated units/modules may be stored in a computer readable memory if implemented in the form of software program modules and sold or used as a stand-alone product. Based on this understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in whole or in part in the form of a software product stored in a memory, comprising several instructions for causing a computer device (which may be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method of the various embodiments of the present application. And the aforementioned memory includes: a usb disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing program codes.

In the foregoing embodiments, the descriptions of the embodiments are emphasized, and for parts of one embodiment that are not described in detail, reference may be made to related descriptions of other embodiments. The technical features of the foregoing embodiments may be arbitrarily combined, and for brevity, all of the possible combinations of the technical features of the foregoing embodiments are not described, however, all of the combinations of the technical features should be considered as being within the scope of the disclosure.

Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the method embodiments described above may be performed by hardware associated with program instructions. The foregoing program may be stored in any computer readable storage medium. The program, when executed, performs steps including the method embodiments described above; and the aforementioned storage medium includes: various media that can store program code, such as ROM, RAM, magnetic or optical disks.

Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the application.

Claims

1. A knowledge graph processing method, characterized in that the method comprises:

2. The method according to claim 1, wherein the constructing the knowledge-graph of the target business according to the data related to the target business by the plurality of objects in the banking system and the information of the target business by adopting a federal knowledge-graph construction method includes:

3. The method of claim 1, wherein the constructing the entity weight set of the sub-service based on the updated knowledge-graph of the target service comprises:

4. A method according to any one of claims 1-3, wherein after said constructing a knowledge-graph of said target business, said method comprises:

5. The target service credit processing method is characterized by comprising the following steps:

Acquiring an entity weight set of the target object from the entity weight set of the target sub-service based on the entity in the knowledge graph of the target object aiming at the target sub-service; the knowledge graph and entity weight set of the target object for the target sub-service are obtained based on the method according to any one of claims 1-4;

6. A knowledge-graph processing apparatus, comprising:

7. A target service trust processing apparatus, comprising:

The acquisition weight set module is used for acquiring the entity weight set of the target object from the entity weight set of the target sub-service based on the entity in the knowledge graph of the target object for the target sub-service; the knowledge graph and entity weight set of the target object for the target sub-service are obtained based on the method according to any one of claims 1-4;

8. An electronic device, comprising: a processor, and a memory communicatively coupled to the processor;

the memory stores computer-executable instructions;

The processor executes computer-executable instructions stored in the memory to implement the method of any one of claims 1-5.

9. A computer readable storage medium having stored therein computer executable instructions which when executed by a processor are adapted to carry out the method of any one of claims 1-5.

10. A computer program product comprising a computer program which, when executed by a processor, implements the method of any of claims 1-5.