Business reasoning method and device, equipment, medium and program product

Abstract

The invention discloses a business reasoning method, device and equipment, a medium and a program product. The invention relates to the technical field of computers. The method comprises the following steps: in response to a service reasoning request initiated by a user side, obtaining a target entity identifier contained in the service reasoning request; searching a target entity network matched with the target entity identifier from the knowledge graph instance according to the target entity identifier; loading a predefined SWRL inference rule set, and adopting an inference engine to execute logical inference operation in the target entity network based on the SWRL inference rule set to generate an inference result associated with the target entity identifier; optimizing and updating the knowledge graph instance based on the new entity attribute and the new entity relationship contained in the reasoning result; and a reasoning result is fed back to the user side through the application programming interface. According to the technical scheme of the invention, the accuracy and timeliness of business reasoning decision can be improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of computer technology, and in particular to a business reasoning method, apparatus, device, medium and program product. Background Technology

[0002] In modern financial services, the business reasoning process is a core business link, and the quality of its decision-making directly affects the risk and efficiency of financial institutions.

[0003] Business reasoning currently faces numerous challenges, particularly in information processing and decision support. Existing technical solutions often lack effective data integration capabilities, failing to effectively integrate multi-source heterogeneous data, leading to information silos and limiting the comprehensiveness and accuracy of the information upon which decisions are based.

[0004] In addition, existing business reasoning relies on traditional rule engines and data analysis methods, which have certain limitations when dealing with complex business reasoning. Their reasoning capabilities are relatively limited and cannot meet user needs. Summary of the Invention

[0005] This invention provides a business reasoning method, apparatus, device, medium, and program product to address the problem of limited information dimensions and reasoning capabilities in the decision-making process.

[0006] In a first aspect, embodiments of the present invention provide a business reasoning method, the method comprising:

[0007] In response to a business reasoning request initiated by the user, obtain the target entity identifier contained in the business reasoning request;

[0008] Based on the target entity identifier, a target entity network matching the target entity identifier is retrieved from the knowledge graph instance; the knowledge graph instance is constructed by mapping and instantiating business data based on a pre-built knowledge graph ontology; the knowledge graph instance contains entities, entity attributes, and entity relationships.

[0009] Load the predefined Semantic Web Rule Language (SWRL) inference rule set, and use the inference engine to perform logical inference operations in the target entity network based on the SWRL inference rule set to generate inference results associated with the target entity identifier;

[0010] Based on the new entity attributes and new entity relationships contained in the reasoning results, the knowledge graph instance is optimized and updated.

[0011] The inference results are fed back to the user terminal via an application programming interface.

[0012] Secondly, embodiments of the present invention also provide a business inference apparatus, the apparatus comprising:

[0013] The entity identifier acquisition module is used to acquire the target entity identifier contained in the business reasoning request in response to the business reasoning request initiated by the user terminal.

[0014] The entity network determination module is used to retrieve a target entity network that matches the target entity identifier from a knowledge graph instance based on the target entity identifier; the knowledge graph instance is constructed by mapping and instantiating business data based on a pre-built knowledge graph ontology; the knowledge graph instance contains entities, entity attributes, and entity relationships.

[0015] The logical reasoning module is used to load a predefined Semantic Web Rule Language (SWRL) reasoning rule set, and use a reasoning engine to perform logical reasoning operations in the target entity network based on the SWRL reasoning rule set to generate reasoning results associated with the target entity identifier.

[0016] The knowledge graph optimization module is used to optimize and update the knowledge graph instance based on the new entity attributes and new entity relationships contained in the reasoning results.

[0017] The inference result feedback module is used to feed back the inference result to the user terminal through the application programming interface.

[0018] Thirdly, embodiments of the present invention also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the business reasoning method as described in any of the embodiments of the present invention.

[0019] Fourthly, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the business reasoning method as described in any of the embodiments of the present invention.

[0020] Fifthly, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the business reasoning method as described in any of the embodiments of the present invention.

[0021] In this embodiment of the invention, in response to a business reasoning request initiated by a user, the target entity identifier contained in the business reasoning request is obtained. Based on the target entity identifier, a target entity network matching the target entity identifier is retrieved from a knowledge graph instance. The knowledge graph instance contains entities, entity attributes, and entity relationships. Then, a predefined SWRL reasoning rule set is loaded, and a reasoning engine is used to perform logical reasoning operations in the target entity network based on the SWRL reasoning rule set, generating a reasoning result associated with the target entity identifier. Based on the new entity attributes and new entity relationships contained in the reasoning result, the knowledge graph instance is optimized and updated, and the reasoning result is fed back to the user through an application programming interface. The knowledge graph instance can integrate multi-source heterogeneous information related to the business to be reasoned. Using the knowledge graph instance as the basis for business reasoning can improve the accuracy and timeliness of decision-making. Furthermore, using the SWRL reasoning rule set for business reasoning can achieve flexible management of reasoning rules and improve reasoning capabilities. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A flowchart of a business reasoning method provided in an embodiment of the present invention;

[0024] Figure 2 A flowchart of another business reasoning method provided in an embodiment of the present invention;

[0025] Figure 3a A flowchart of yet another business reasoning method provided in an embodiment of the present invention;

[0026] Figure 3b This is a data mapping diagram provided for an embodiment of the present invention;

[0027] Figure 4 This is a schematic diagram of the structure of a business inference device provided in an embodiment of the present invention;

[0028] Figure 5 This is a schematic diagram of the structure of a business inference electronic device provided in an embodiment of the present invention. Detailed Implementation

[0029] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. In addition, it should be noted that for the convenience of description, only the parts related to the present invention rather than all the structures are shown in the accompanying drawings.

[0030] It should be noted that: similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance. The acquisition, storage, use, processing, etc. of data in the technical solution of this application all comply with the relevant regulations of national laws and regulations. It should be noted that in the embodiments of this application, some industry-existing solutions such as certain software, components, models, etc. may be mentioned. They should be regarded as exemplary, and their purpose is only to illustrate the feasibility in the implementation of the technical solution of this application, but it does not mean that the applicant has already or necessarily used this solution.

[0031] In modern financial services, the business reasoning process is the core business link, and its decision-making quality directly relates to the risks and benefits of financial institutions. In some scenarios of business handling qualification approval, it is necessary to comprehensively consider multi-dimensional information of users to determine whether users have the qualification to handle. For example, in the personal loan approval business, it is necessary to process and analyze data from different sources and in different formats to comprehensively evaluate whether users can pass the loan approval process.

[0032] Existing technical solutions often lack effective data integration capabilities and cannot effectively integrate multi-source heterogeneous data, resulting in the phenomenon of information silos, thus affecting the accuracy and timeliness of the final approval decision.

[0033] In this embodiment of the invention, in response to a business reasoning request initiated by a user, the target entity identifier contained in the business reasoning request is obtained. Based on the target entity identifier, a target entity network matching the target entity identifier is retrieved from a knowledge graph instance. The knowledge graph instance contains entities, entity attributes, and entity relationships. Then, a predefined SWRL reasoning rule set is loaded, and a reasoning engine is used to perform logical reasoning operations in the target entity network based on the SWRL reasoning rule set, generating a reasoning result associated with the target entity identifier. Based on the new entity attributes and new entity relationships contained in the reasoning result, the knowledge graph instance is optimized and updated, and the reasoning result is fed back to the user through an application programming interface. The knowledge graph instance can integrate multi-source heterogeneous information related to the business to be reasoned. Using the knowledge graph instance as the basis for business reasoning can improve the accuracy and timeliness of decision-making. Furthermore, using the SWRL reasoning rule set for business reasoning can achieve flexible management of reasoning rules and improve reasoning capabilities.

[0034] Figure 1 The flowchart of the business reasoning method provided in this embodiment of the invention is applicable to situations where business reasoning is performed based on an indicator graph. This method can be executed by a business reasoning device, which can be implemented in hardware and / or software. For example... Figure 1 As shown, the method includes the following steps:

[0035] S101. In response to a business reasoning request initiated by the user, obtain the target entity identifier contained in the business reasoning request.

[0036] A business reasoning request is a machine-readable instruction initiated by the user client to perform business reasoning and decision-making on specific business objects within a knowledge graph instance. Its fundamental purpose is to trigger logical reasoning operations based on the knowledge graph instance and rule engine. The target entity identifier is the core parameter of the business reasoning request. The target entity identifier is a unique identifier for a specific entity within the knowledge graph instance and serves as the entry point for the business reasoning operation.

[0037] In addition to the target entity identifier, the business reasoning request may also include other parameters to limit the scope of reasoning, such as limiting the specific rule set used in this reasoning or limiting the business scope of this reasoning.

[0038] For example, a business reasoning request is a loan approval request, which includes a target entity identifier such as a loan applicant identifier or a loan business identifier.

[0039] In this embodiment of the invention, after receiving a business reasoning request submitted by the user, the business reasoning request is parsed to obtain the target entity identifier contained in the business reasoning request, so as to subsequently determine the target entity to be reasoned in the knowledge graph instance based on the target entity identifier. For example, after receiving a loan approval request submitted by the user, the loan approval request is parsed to obtain the loan applicant identifier or loan business identifier contained therein.

[0040] S102. Based on the target entity identifier, retrieve the target entity network that matches the target entity identifier from the knowledge graph instance.

[0041] A knowledge graph instance is constructed by mapping and instantiating business data based on a pre-built knowledge graph ontology; the knowledge graph instance contains entities, entity attributes, and entity relationships.

[0042] A knowledge graph ontology is an explicit specification of concepts and relationships between concepts in a specific domain, forming the architecture for subsequent construction of knowledge graph instances. For example, a knowledge graph ontology is constructed using the Protégé tool based on OWL (WebOntology Language).

[0043] Ontology classes and attributes constitute the basic elements of a knowledge graph ontology. An ontology class is a conceptual identifier, while attributes describe specific aspects of an ontology class. OWL is used to semantically describe entity classes. It uses a formal language to define ontology classes, attributes, and relationships between ontology classes, offering advantages such as clear structure, strong formalization, and high machine interpretability.

[0044] A knowledge graph instance is an instantiated knowledge graph that uses a knowledge graph ontology as a framework to map business data to the corresponding ontology classes and attributes in the knowledge graph ontology. The knowledge graph instance contains entities, entity attributes, and entity relationships.

[0045] In this embodiment of the invention, after parsing the target entity identifier, the target entity network corresponding to the target entity identifier is retrieved from the knowledge graph instance. Specifically, the target entity identifier is used as the search key to query the constructed knowledge graph instance. First, the target entity corresponding to the target entity identifier is located, for example, a specific loan applicant, Zhang San. Then, with this target entity as the center, according to a preset traversal depth, such as first-degree or second-degree relationships, all related entities directly or indirectly connected to the target entity through entity relationships are located and retrieved, such as Zhang San's loan application, his collateral, and related guarantors. This forms a target entity network centered on the target entity and associated with multiple related entities, i.e., a subgraph of the knowledge graph instance serves as the target entity network.

[0046] In a specific example, based on the loan applicant's identifier in the loan approval request initiated by the user, the target entity corresponding to the loan applicant's identifier is located from the knowledge graph instance. Then, with the target entity as the center, according to the preset traversal depth, such as 3 degrees, the related entities that are directly or indirectly related to the target entity are searched, and the subgraph of the knowledge graph instance with a depth of 3 is obtained as the target entity network.

[0047] S103. Load the predefined Semantic Web Rule Language (SWRL) inference rule set, and use the inference engine to perform logical inference operations in the target entity network based on the SWRL inference rule set to generate inference results associated with the target entity identifier.

[0048] The Semantic Web Rule Language (SWRL) inference rule set is a formalized set of rules consisting of one or more SWRL rules used for logical reasoning on a knowledge graph. Each SWRL rule consists of two parts: premises and conclusion. The premises describe a set of conditions that must be met, which are formed by a series of atomic expressions connected by logical operators such as AND, OR, and NOT, defining the conditions that trigger the rule. The conclusion describes the action to be performed or the fact to be inferred when the premises of the rule are met.

[0049] A pre-defined set of SWRL inference rules guides the reasoning process by constructing multiple rigorously defined logical rules. New facts or relationships are derived based on existing knowledge, and potential entity relationships or attributes are predicted to help fill in missing information, improve data integrity, and provide logical support for complex decision-making.

[0050] In this embodiment of the invention, a predefined Semantic Web Rule Language (SWRL) inference rule set is first loaded. Then, an inference engine, such as Apache Jena, is used to match these formalized logical rules with entities and entity attributes in the target entity network to determine whether the preconditions in each SWRL inference rule can be satisfied in the target entity network. If they are satisfied, the conclusion corresponding to the precondition is triggered, that is, the conclusion is used as the inference result associated with the target entity identifier.

[0051] Because SWRL rules are explicitly declared, they allow for tracing how reasoning derives specific conclusions from a set of premises, resulting in highly interpretable reasoning processes and outcomes. Furthermore, the SWRL inference rule set exists in a logical form close to natural language, decoupled from program code. Therefore, it can be flexibly added, modified, or deleted as needed, enabling inference operations to adapt to changing knowledge and different requirements.

[0052] In a specific example, one SWRL rule in the SWRL inference rule set is: Premise: The value of the collateral provided by the loan applicant is greater than the loan amount requested; Conclusion: Loan approved. This logical reasoning is then performed within the target entity network associated with the loan applicant to determine whether the value of the collateral provided by the loan applicant exceeds the loan amount, thereby arriving at the inference conclusion.

[0053] S104. Based on the new entity attributes and new entity relationships contained in the reasoning results, optimize and update the knowledge graph instances.

[0054] In this embodiment of the invention, the new entity attributes and new entity relationships generated from the reasoning are converted into a format acceptable to the knowledge graph instance, such as Resource Description Framework (RDF) triples, and then written back to the knowledge graph instance through a data interface. By feeding the reasoning results back into the knowledge graph instance, the knowledge graph instance is no longer a static knowledge base, but can be updated in real time based on the reasoning results, serving as a new basis for subsequent decisions. For example, an "Approval Status: Approved" attribute is added to the "Loan Application" entity.

[0055] S105. Feed back the inference results to the user terminal through the application programming interface.

[0056] In this embodiment of the invention, the reasoning result, such as whether a loan application is approved, is fed back to the user's terminal via an application programming interface (API) as the basis for the user's final decision. Simultaneously with the feedback of the reasoning result, the corresponding reasoning details can also be fed back to the user's terminal. By providing detailed reasoning evidence, an interpretable basis for the user's decision can be offered.

[0057] The technical solution of this invention responds to a business reasoning request initiated by a user terminal, obtains the target entity identifier contained in the business reasoning request, retrieves the target entity network matching the target entity identifier from a knowledge graph instance based on the target entity identifier, wherein the knowledge graph instance contains entities, entity attributes, and entity relationships, then loads a predefined SWRL reasoning rule set, and uses a reasoning engine to perform logical reasoning operations in the target entity network based on the SWRL reasoning rule set, generating a reasoning result associated with the target entity identifier, optimizing and updating the knowledge graph instance based on the new entity attributes and new entity relationships contained in the reasoning result, and feeding back the reasoning result to the user terminal through an application programming interface. The knowledge graph instance can integrate multi-source heterogeneous information related to the business to be reasoned. Using the knowledge graph instance as the basis for business reasoning can improve the accuracy and timeliness of decision-making, and using the SWRL reasoning rule set for business reasoning can achieve flexible management of reasoning rules and improve reasoning capabilities.

[0058] Figure 2 This is a flowchart of a business reasoning method provided in an embodiment of the present invention. Based on the above embodiments, this embodiment provides specific steps for retrieving a target entity network matching the target entity identifier from a knowledge graph instance, and for performing logical reasoning operations in the target entity network using a reasoning engine based on the SWRL reasoning rule set to generate a reasoning result associated with the target entity identifier. Figure 2 As shown, the method includes the following steps:

[0059] S201. In response to a business reasoning request initiated by the user, obtain the target entity identifier contained in the business reasoning request.

[0060] S202. Identify the target entity corresponding to the target entity identifier in the knowledge graph instance.

[0061] A knowledge graph instance is constructed by mapping and instantiating business data based on a pre-built knowledge graph ontology; the knowledge graph instance contains entities, entity attributes, and entity relationships.

[0062] In this embodiment of the invention, the target entity corresponding to the target entity identifier is queried in the knowledge graph instance based on the target entity identifier. For example, the corresponding loan applicant is found in the knowledge graph instance based on the loan applicant identifier contained in the loan approval request.

[0063] S203. Using the target entity as the root node, traverse the associated entities connected to the target entity through entity relationships to obtain a subgraph of the knowledge graph instance with a preset network depth, which serves as the target entity network.

[0064] In this embodiment of the invention, a knowledge graph instance traversal algorithm, such as a depth-first or breadth-first algorithm, is initiated with the target entity as the root node. This traverses multiple associated entities that are directly or indirectly connected to the target entity through entity relationships, thereby obtaining a subgraph of the knowledge graph instance as the target entity network. The traversal scope can be determined based on the network depth included in the business reasoning request. Information scattered across different data sources and associated with the target entity is aggregated into a target entity network linked by entity relationships, providing a highly integrated data foundation for subsequent logical reasoning and thus improving the reliability of the reasoning results. Furthermore, a controllable balance between reasoning performance and information completeness can be achieved through flexible settings of the network depth.

[0065] For example, if the network depth in a business reasoning request is 1, then only the associated entities and entity relationships directly connected to the root node are traversed; if the network depth is 2, then the associated entities directly connected to the root node are first-degree associated entities, and then the associated entities directly connected to the first-degree entities are second-degree entities. Finally, all the traversed entities, relationships, and attributes are associated to form a subgraph of the knowledge graph instance, which is the target entity network.

[0066] S204. Load the predefined SWRL inference rule set and use the inference engine to extract the entities to be inferred contained in the SWRL inference rule set, as well as the attributes of the entities to be inferred associated with the entities to be inferred, from the target entity network.

[0067] In this embodiment of the invention, after determining the target entity network, a predefined SWRL inference rule set is read from persistent storage and parsed into memory objects operable within the inference engine. Then, the inference engine matches each SWRL rule in the SWRL inference rule set with the target entities and entity attributes in the target entity network to obtain the entities to be inferred that match the premise of the SWRL rule, as well as the entity attributes associated with those entities.

[0068] In a specific example, the premise of the SWRL rule is: the loan applicant has an outstanding delinquency record, and the corresponding conclusion is: the loan application is rejected. An inference engine can be used to extract the entities to be inferred from the target entity network, including the loan applicant and the delinquency record. The attributes associated with these entities include information related to the loan applicant's employment status, and information related to the delinquency record such as the delinquency date, delinquency amount, and delinquency status.

[0069] Optional, also includes:

[0070] Monitor whether the entity attributes of each entity in the knowledge graph instance have changed;

[0071] When a change in the entity attribute of a target entity is detected, the inference engine is used to perform logical inference operations in the target entity network associated with the target entity based on the SWRL inference rule set.

[0072] In this optional embodiment, a scheduled task can be deployed to periodically scan whether the entity attributes of each entity in the knowledge graph instance have changed. Alternatively, a hook function can be set at the write application interface level of the knowledge graph instance; when an entity attribute changes, the hook function will be triggered immediately, outputting the change information.

[0073] Furthermore, when a change in the entity attributes of a target entity is detected, the inference engine, based on SWRL inference rules, performs logical inference within the target entity network associated with the target entity. The specific method for determining the target entity network associated with the target entity is the same as in S203 and will not be repeated here. By proactively sensing changes in knowledge graph instances, logical inference operations can be re-executed promptly based on the changed entity attributes, improving the timeliness of business inference. In financial risk control scenarios, this can reduce risk and avoid potential losses.

[0074] S205. Based on the logical judgment conditions in the SWRL inference rule set, perform condition judgments on the attributes of the entity to be inferred, and obtain the inference result associated with the target entity identifier.

[0075] In this embodiment of the invention, after determining the entity to be inferred and its attributes, the attributes of the entity to be inferred are conditionally judged according to the logical judgment conditions in the SWRL inference rule set. That is, it is judged whether the attributes of the entity to be inferred can satisfy the premise part in the SWRL rule. Finally, based on the judgment result, the inference result associated with the target entity identifier is determined.

[0076] By combining the SWRL inference rule set with a highly integrated target entity network of multi-source data, complex business reasoning that relies on human experience can be transformed into an automated reasoning process that uses an inference engine, is based on explicit inference rules, and is grounded in multi-source data. This improves decision-making efficiency and the reliability of inference results.

[0077] In a specific example, the premise of the SWRL rule is: the loan applicant has an outstanding delinquency record, and the corresponding conclusion is: the loan application is rejected. After the inference engine extracts the loan applicant, delinquency record, and related entity attributes such as employment status, delinquency date, delinquency amount, and delinquency status from the target entity network, it further determines whether the loan applicant has a delinquency record among the aforementioned entity attributes. If so, the premise of the SWRL rule is satisfied, and the inference result is that the loan application is rejected.

[0078] Optionally, after performing logical reasoning operations in the target entity network and generating a reasoning result associated with the target entity identifier, the method further includes:

[0079] Record the historical records of each SWRL rule being triggered and the inference results in the SWRL inference rule set;

[0080] Based on the historical records and user feedback on the inference results, the performance index of each SWRL inference rule is calculated.

[0081] If the performance index is lower than the set index threshold, generate rule optimization prompts or disable rule prompts.

[0082] In this optional embodiment, specific steps are provided after performing logical reasoning operations in the target entity network and generating a reasoning result associated with the target entity identifier: when executing SWRL rules in the SWRL reasoning rule set, the time when each SWRL rule is triggered, which specific entities and entity attributes this trigger is based on, and the final output reasoning result are recorded.

[0083] Subsequently, user feedback on each reasoning result on the approval page can be obtained, including "Agree / Reject," with "Agree" considered positive feedback and "Reject" considered negative feedback. The reasoning results can also be compared with the final business facts to obtain feedback information for each reasoning result.

[0084] Furthermore, the performance metrics of each SWRL rule in the SWRL inference rule set are statistically analyzed. For example, the inference accuracy is obtained as the performance metric for each SWRL rule by calculating the ratio of the number of times the inference result is correct to the number of times the rule is triggered. Finally, if the performance metric falls below a set threshold, rule optimization suggestions are generated or the suggestion is disabled. By calculating performance metrics such as accuracy and conflict rate, the business value and health of each SWRL rule are measured, allowing for timely intervention in SWRL rules with logical flaws, further improving the reliability and stability of business inference.

[0085] S206. Based on the new entity attributes and new entity relationships contained in the reasoning results, optimize and update the knowledge graph instances.

[0086] S207. Feed back the inference results to the user terminal through the application programming interface.

[0087] The technical solution of this invention responds to a business reasoning request initiated by a user terminal, obtains the target entity identifier contained in the business reasoning request, determines the target entity corresponding to the target entity identifier in the knowledge graph instance, uses the target entity as the root node, traverses the associated entities connected to the target entity through entity relationships to obtain a subgraph of the knowledge graph instance with a preset network depth, which serves as the target entity network, loads a predefined SWRL reasoning rule set, and uses a reasoning engine to extract the entities to be reasoned contained in the SWRL reasoning rule set and the attributes of the entities to be reasoned associated with the entities to be reasoned from the target entity network. According to the logical judgment conditions in the SWRL reasoning rule set, condition judgments are performed on the attributes of the entities to be reasoned to obtain the reasoning result associated with the target entity identifier. Based on the new entity attributes and new entity relationships contained in the reasoning result, the knowledge graph instance is optimized and updated, and the reasoning result is fed back to the user terminal through an application programming interface. The information scattered across different data sources associated with the target entity is aggregated into a target entity network associated with entity relationships, providing a highly integrated data foundation for subsequent logical reasoning, thereby improving the reliability of the reasoning result. Furthermore, when determining the target entity network, a controllable balance between reasoning performance and information completeness can be achieved through flexible settings of the network depth.

[0088] Figure 3a This is a flowchart of a business reasoning method provided in an embodiment of the present invention. Based on the above embodiments, this embodiment provides specific steps before obtaining the target entity identifier contained in the business reasoning request in response to a business reasoning request initiated by the user terminal, and specific steps for feeding back the reasoning result to the user terminal through an application programming interface. Figure 3a As shown, the method includes the following steps:

[0089] S301. Obtain the original business data corresponding to the ontology classes, class attributes, and class relationships in the knowledge graph ontology from at least one business data source.

[0090] The knowledge graph ontology is constructed using the Web Ontology Language and includes multiple ontology classes, class attributes, and class relationships.

[0091] In this embodiment of the invention, the original business data corresponding to the ontology classes, class attributes, and class relationships in the knowledge graph ontology is obtained from at least one business data source through an external interface module. The business data source may include a structured data source storing structured data, such as a relational database; a semi-structured data source storing semi-structured data, such as a NoSQL database; and an unstructured data source such as contract documents.

[0092] The process involves extracting raw business data corresponding to the ontology classes, class attributes, and class relationships in the knowledge graph ontology from the aforementioned multiple business data sources. The specific steps are as follows: Figure 3b As shown, we can first analyze the ontology structure and database structure of the knowledge graph to determine the fields corresponding to ontology classes, class attributes, and class relationships within the same ontology structure. Then, we can write SQL queries to extract raw business data related to the ontology classes, class attributes, and class relationships from various data sources. For example, for the ontology class "loan applicant," we can extract loan applicant information from the customer information table and, by associating it with the customer identifier in the loan application master table, extract their loan records, etc.

[0093] S302. According to the preset mapping rules, the original business data is instantiated into entities of the corresponding ontology class, and the entity attributes and entity relationships corresponding to the entities are determined to obtain knowledge graph instances.

[0094] Mapping rules are used to associate ontology classes with corresponding actual business data, defining entity mapping, attribute mapping, and relationship mapping. For example, in the ontology class, "Loan Applicant" corresponds to the 'CUSTOMER_ID' field in the "Customer Information Table"; in the class attribute, "Annual Income" corresponds to the 'Income' field in the "Customer Information Table"; when the 'CUSTOMER_ID' in the "Loan Application Table" is equal to the 'ID' in the "Customer Information Table," a loan application relationship is established between the two corresponding entities.

[0095] In this embodiment of the invention, based on entity mapping in the preset mapping rules, actual business data from different data sources are instantiated into entities of corresponding ontology classes. Entity attributes and relationships corresponding to the entities are determined based on attribute mapping and relationship mapping, resulting in the final knowledge graph instance. By integrating data from different data sources through mapping rules, a knowledge graph instance is obtained, providing data support for subsequent reasoning. Furthermore, when the business data source structure changes, only the mapping rules need to be adjusted to achieve data mapping without modifying the core code, allowing the knowledge graph instance to flexibly adapt to business changes.

[0096] Optionally, the technical solution of this embodiment of the invention further includes:

[0097] In response to a user's update operation on the knowledge graph ontology, obtain the target ontology class corresponding to the update operation;

[0098] According to the preset mapping rules, obtain the target business data associated with the target ontology class from at least one business data source;

[0099] The target business data is instantiated into a target entity of the corresponding target ontology class, and the entity attributes and entity relationships corresponding to the target entity are determined to obtain the updated knowledge graph instance.

[0100] In this optional embodiment, after receiving an update operation from a user regarding the knowledge graph ontology, the target ontology class corresponding to the update operation can be obtained. The update operation includes adding, deleting, or modifying ontology classes within the knowledge graph ontology. Then, according to preset mapping rules, target business data associated with the updated target ontology class is obtained from at least one business data source. The target business data is then instantiated into a target entity corresponding to the target ontology class, and the entity attributes and relationships corresponding to the target entity are determined, resulting in an updated knowledge graph instance. Through the corresponding update of the knowledge graph instance, dynamic collaborative evolution between the knowledge graph schema layer and data layer is achieved, eliminating the need for manual modification of the knowledge graph instance, reducing maintenance costs, and ensuring the reliability of the inference results.

[0101] Optionally, after obtaining the knowledge graph instance, it also includes:

[0102] The knowledge graph instances are stored in the form of Resource Description Framework (RDF) files.

[0103] In this optional embodiment, after obtaining the knowledge graph instance, the knowledge graph instance is further stored in the form of a Resource Description Framework (RDF) file to ensure efficient utilization of the knowledge graph instance.

[0104] S303. In response to a business reasoning request initiated by the user, obtain the target entity identifier contained in the business reasoning request.

[0105] Optionally, the business reasoning request is a loan business request; the ontology class includes at least one of loan applicant, loan application, collateral, guarantor, and credit score; the class relationship includes at least one of the loan application relationship between the loan applicant and the loan application, and the collateral relationship between the loan applicant and the collateral.

[0106] Optionally, the target entity identifier may include a loan applicant identifier or a loan business identifier.

[0107] In the two optional embodiments described above, the business reasoning request is defined as a loan business request. Adaptively, the ontology class includes at least one of the following: loan applicant, loan application, collateral, guarantor, and credit score. The class relationship includes at least one of the following: the loan application relationship between the loan applicant and the loan application, and the collateral relationship between the loan applicant and the collateral.

[0108] When a user initiates a business reasoning request, the target entity identifier included can be either the loan applicant's identifier or the loan business identifier.

[0109] S304. Based on the target entity identifier, retrieve the target entity network that matches the target entity identifier from the knowledge graph instance.

[0110] The knowledge graph instance is constructed by mapping and instantiating business data based on a pre-built knowledge graph ontology; the knowledge graph instance contains entities, entity attributes, and entity relationships.

[0111] S305. Load the predefined Semantic Web Rule Language (SWRL) inference rule set, and use the inference engine to perform logical inference operations in the target entity network based on the SWRL inference rule set to generate inference results associated with the target entity identifier.

[0112] S306. Based on the new entity attributes and new entity relationships contained in the reasoning results, optimize and update the knowledge graph instances.

[0113] Optionally, based on the new entity attributes and new entity relationships contained in the reasoning result, the knowledge graph instance is optimized and updated, including at least one of the following:

[0114] Based on the new entity attributes in the reasoning results, add new entity attributes to existing entities, or modify the entity attributes of existing entities.

[0115] Based on the new entity relationships in the reasoning results, a new entity relationship is established between two existing entities.

[0116] In this optional embodiment, a specific method is provided for optimizing and updating the knowledge graph instance based on the new entity attributes and new entity relationships contained in the reasoning result: This can be achieved by adding new entity attributes to existing entities or modifying the entity attributes of existing entities based on the new entity attributes in the reasoning result. For example, regarding a user's loan approval request (including the loan applicant's statement), if the reasoning result is "agree," then the entity attribute "agree to loan" can be added to the entity corresponding to the loan applicant in the knowledge graph instance.

[0117] In addition, new entity relationships can be established between two existing entities based on the new entity relationships in the reasoning results. For example, if the reasoning result can assert a relationship, such as "customer Zhang San and customer Li Si have associated risks," then an association relationship can be established between the two existing entities in the knowledge graph instance. By feeding the reasoning results back to the knowledge graph instance, the knowledge graph can learn new knowledge in each reasoning process, serving as the basis for subsequent business reasoning, thereby continuously optimizing business reasoning and improving the reliability of subsequent decisions.

[0118] S307. The reasoning details of performing logical reasoning operations in the target entity network based on the SWRL reasoning rule set are fed back to the user end through the application programming interface.

[0119] The reasoning details include the reasoning process and reasoning results for each reasoning rule in the SWRL reasoning rule set.

[0120] In this embodiment of the invention, to improve the interpretability of the inference results, the inference details, based on the SWRL inference rule set, performed logical inference operations in the target entity network, are fed back to the user terminal via an application programming interface. The inference details may include the SWRL rules upon which the inference result is based, the premises of the SWRL rules, the inference process, and the corresponding inference result. By feeding back the inference details to the user terminal, when the user needs to verify the inference result, there is no need to search through data in multiple systems; instead, the user can directly obtain the SWRL rules upon which the inference result is based and the corresponding inference process.

[0121] The technical solution of this invention involves obtaining original business data corresponding to ontology classes, class attributes, and class relationships in a knowledge graph ontology from at least one business data source. Based on preset mapping rules, the original business data is instantiated into entities corresponding to the ontology classes, and the entity attributes and entity relationships corresponding to the entities are determined to obtain a knowledge graph instance. In response to a business reasoning request initiated by a user, the target entity identifier contained in the business reasoning request is obtained. Based on the target entity identifier, a target entity network matching the target entity identifier is retrieved from the knowledge graph instance. A predefined Semantic Web Rule Language (SWRL) reasoning rule set is loaded, and a reasoning engine is used to perform logical reasoning operations in the target entity network based on the SWRL reasoning rule set, generating a reasoning result associated with the target entity identifier. Based on the new... The entity attributes and new entity relationships are used to optimize and update the knowledge graph instance. The reasoning details, based on the SWRL reasoning rule set, are fed back to the user end through the application programming interface. On the one hand, the mapping rules integrate data from different data sources to obtain the knowledge graph instance, providing data basis for subsequent reasoning. When the business data source structure changes, only the mapping rules need to be adjusted to achieve data mapping, without modifying the core code, so that the knowledge graph instance can flexibly adapt to business changes. On the other hand, by feeding back the reasoning details to the user end, when the user needs to verify the reasoning results, there is no need to search through multiple systems for data. Instead, the user can intuitively obtain the SWRL rules on which the reasoning results are based and the corresponding reasoning process.

[0122] Figure 4 This is a schematic diagram of the structure of a business inference device provided in an embodiment of the present invention. Figure 4 As shown, the business reasoning device includes: an entity identifier acquisition module 401, an entity network determination module 402, a logical reasoning module 403, a knowledge graph optimization module 404, and a reasoning result feedback module 405.

[0123] The entity identifier acquisition module 401 is used to acquire the target entity identifier contained in the business reasoning request in response to the business reasoning request initiated by the user terminal.

[0124] The entity network determination module 402 is used to retrieve a target entity network that matches the target entity identifier from a knowledge graph instance based on the target entity identifier; the knowledge graph instance is constructed by mapping and instantiating business data based on a pre-built knowledge graph ontology; the knowledge graph instance contains entities, entity attributes, and entity relationships.

[0125] The logical reasoning module 403 is used to load a predefined Semantic Web Rule Language (SWRL) reasoning rule set, and use a reasoning engine to perform logical reasoning operations in the target entity network based on the SWRL reasoning rule set to generate reasoning results associated with the target entity identifier.

[0126] The knowledge graph optimization module 404 is used to optimize and update the knowledge graph instance based on the new entity attributes and new entity relationships contained in the reasoning result.

[0127] The reasoning result feedback module 405 is used to feed back the reasoning result to the user terminal through the application programming interface.

[0128] The technical solution of this invention responds to a business reasoning request initiated by a user terminal, obtains the target entity identifier contained in the business reasoning request, retrieves the target entity network matching the target entity identifier from a knowledge graph instance based on the target entity identifier, wherein the knowledge graph instance contains entities, entity attributes, and entity relationships, then loads a predefined SWRL reasoning rule set, and uses a reasoning engine to perform logical reasoning operations in the target entity network based on the SWRL reasoning rule set, generating a reasoning result associated with the target entity identifier, optimizing and updating the knowledge graph instance based on the new entity attributes and new entity relationships contained in the reasoning result, and feeding back the reasoning result to the user terminal through an application programming interface. The knowledge graph instance can integrate multi-source heterogeneous information related to the business to be reasoned. Using the knowledge graph instance as the basis for business reasoning can improve the accuracy and timeliness of decision-making, and using the SWRL reasoning rule set for business reasoning can achieve flexible management of reasoning rules and improve reasoning capabilities.

[0129] Optionally, the entity network determination module 402 is specifically used for:

[0130] In the knowledge graph instance, identify the target entity corresponding to the target entity identifier;

[0131] Using the target entity as the root node, traverse the associated entities connected to the target entity through entity relationships to obtain a subgraph of the knowledge graph instance with a preset network depth, which serves as the target entity network.

[0132] Optional, logic reasoning module 403, specifically used for:

[0133] Using an inference engine, extract the entities to be inferred contained in the SWRL inference rule set and the attributes of the entities to be inferred associated with the entities to be inferred from the target entity network;

[0134] Based on the logical judgment conditions in the SWRL reasoning rule set, the attributes of the entity to be reasoned are conditionally judged to obtain the reasoning result associated with the target entity identifier.

[0135] Optionally, the knowledge graph optimization module 404 is used to perform at least one of the following operations:

[0136] Based on the new entity attributes in the reasoning results, add new entity attributes to existing entities, or modify the entity attributes of existing entities.

[0137] Based on the new entity relationships in the reasoning results, a new entity relationship is established between two existing entities.

[0138] Optionally, the knowledge graph ontology is constructed using a web ontology language and includes multiple ontology classes, class attributes, and class relationships.

[0139] The business inference device also includes:

[0140] The raw business data extraction module is used to obtain raw business data corresponding to the ontology class, class attribute and class relationship in the knowledge graph ontology from at least one business data source before obtaining the target entity identifier contained in the business reasoning request initiated by the user terminal.

[0141] The knowledge graph construction module is used to instantiate the original business data into entities of the corresponding ontology class according to preset mapping rules, and determine the entity attributes and entity relationships corresponding to the entities to obtain a knowledge graph instance.

[0142] Optionally, the business inference device also includes:

[0143] The target ontology class determination module is used to obtain the target ontology class corresponding to the update operation in response to the user's update operation on the knowledge graph ontology.

[0144] The target business data extraction module is used to obtain target business data associated with the target ontology class from at least one business data source according to a preset mapping rule.

[0145] The knowledge graph update module is used to instantiate the target business data into a target entity of the corresponding target ontology class, and determine the entity attributes and entity relationships corresponding to the target entity to obtain the updated knowledge graph instance.

[0146] Optional, the inference result feedback module 405 is specifically used for:

[0147] The reasoning details of logical reasoning operations performed in the target entity network based on the SWRL reasoning rule set will be fed back to the user end through the application programming interface.

[0148] The reasoning details include the reasoning process and reasoning results for each reasoning rule in the SWRL reasoning rule set.

[0149] Optionally, the business inference device also includes:

[0150] The entity attribute monitoring module is used to monitor whether the entity attributes of each entity in the knowledge graph instance have changed;

[0151] The inference replay module is used to execute logical inference operations in the target entity network associated with the target entity when a change in the entity attribute of the target entity is detected. This is done using the inference engine and based on the SWRL inference rule set.

[0152] Optionally, the business inference device also includes:

[0153] The history module is used to perform logical reasoning operations in the target entity network, generate reasoning results associated with the target entity identifier, and then record the history of each SWRL rule being triggered and the reasoning results in the SWRL reasoning rule set.

[0154] The performance index calculation module is used to calculate the performance index of each SWRL inference rule based on the historical records and user feedback information on the inference results.

[0155] The rule optimization prompt module is used to generate rule optimization prompts or disable rule prompts when the performance index is lower than a set index threshold.

[0156] Optionally, the business inference device also includes:

[0157] The knowledge graph storage module is used to store the knowledge graph instance in the form of a Resource Description Framework (RDF) file after obtaining the knowledge graph instance.

[0158] Optionally, the business reasoning request is a loan business request; the ontology class includes at least one of loan applicant, loan application, collateral, guarantor, and credit score; the class relationship includes at least one of the loan application relationship between the loan applicant and the loan application, and the collateral relationship between the loan applicant and the collateral.

[0159] Optionally, the target entity identifier may include a loan applicant identifier or a loan business identifier.

[0160] The business reasoning apparatus provided in the embodiments of the present invention can execute the business reasoning method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0161] Figure 5 A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0162] like Figure 5 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0163] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0164] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as business inference methods.

[0165] In some embodiments, the business reasoning method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the business reasoning method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to execute the business reasoning method by any other suitable means (e.g., by means of firmware).

[0166] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0167] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0168] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0169] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0170] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0171] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0172] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the business reasoning method provided in any embodiment of this application.

[0173] In implementing the computer program product, computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0174] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A business reasoning method, characterized by, The method comprises the following steps: In response to a business reasoning request initiated by a user terminal, obtaining a target entity identifier contained in the business reasoning request; According to the target entity identifier, a target entity network matching the target entity identifier is retrieved from a knowledge graph instance; the knowledge graph instance is constructed by mapping and instantiating business data based on a pre-constructed knowledge graph ontology; the knowledge graph instance contains entities, entity attributes, and entity relationships; Loading a predefined semantic web rule language (SWRL) reasoning rule set and using a reasoning engine to perform a logical reasoning operation in the target entity network based on the SWRL reasoning rule set, and generating a reasoning result associated with the target entity identifier; Based on the new entity attributes and new entity relationships contained in the reasoning result, the knowledge graph instance is updated and optimized; The reasoning result is fed back to the user terminal through an application programming interface.

2. The method of claim 1, wherein, According to the target entity identifier, a target entity network matching the target entity identifier is retrieved from a knowledge graph instance, which comprises the following steps: In the knowledge graph instance, determine the target entity corresponding to the target entity identifier; Taking the target entity as the root node, traverse the associated entities connected to the target entity through entity relationships to obtain a subgraph of the knowledge graph instance with a preset network depth as the target entity network.

3. The method of claim 1, wherein, Using a reasoning engine, based on the SWRL reasoning rule set, a logical reasoning operation is performed in the target entity network to generate a reasoning result associated with the target entity identifier, which comprises the following steps: Using a reasoning engine, extract the to-be-reasoned entities and the to-be-reasoned entity attributes associated with the to-be-reasoned entities contained in the SWRL reasoning rule set in the target entity network; According to the logical judgment conditions in the SWRL reasoning rule set, conditionally judge the to-be-reasoned entity attributes to obtain the reasoning result associated with the target entity identifier.

4. The method of claim 1, wherein, Based on the new entity attributes and new entity relationships contained in the reasoning result, the knowledge graph instance is updated and optimized, which comprises at least one of the following: According to the new entity attributes in the reasoning result, add new entity attributes to the existing entities or modify the entity attributes of the existing entities; According to the new entity relationships in the reasoning result, establish new entity relationships between two existing entities.

5. The method of claim 1, wherein, The knowledge graph ontology is constructed by using the web ontology language, which comprises a plurality of ontology classes, class attributes, and class relationships; Before responding to the business reasoning request initiated by the user terminal and obtaining the target entity identifier contained in the business reasoning request, the following steps are further included: Obtain the original business data corresponding to the ontology classes, class attributes, and class relationships in the knowledge graph ontology from at least one business data source; According to the preset mapping rule, the original business data is instantiated as an entity corresponding to the ontology class, and the entity attributes and entity relationships corresponding to the entity are determined to obtain the knowledge graph instance.

6. The method of claim 5, wherein, Further comprising: In response to the user's update operation on the knowledge graph ontology, obtaining a target ontology class corresponding to the update operation; According to a preset mapping rule, target business data associated with the target ontology class is obtained from at least one business data source; The target business data is instantiated as a target entity corresponding to the target ontology class, and entity attributes and entity relationships corresponding to the target entity are determined, to obtain an updated knowledge graph instance.

7. The method of claim 1, wherein, The reasoning result is fed back to the user end through an application programming interface, including: The inference details of performing a logical inference operation in the target entity network based on the SWRL inference rule set are fed back to the user end through an application programming interface; The inference details include an inference process and an inference result for each inference rule in the SWRL inference rule set.

8. The method of claim 1, wherein, Also including: Monitoring whether the entity attributes of each entity in the knowledge graph instance change; When detecting that the entity attributes of a target entity change, performing a logical inference operation in a target entity network associated with the target entity based on the SWRL inference rule set using an inference engine.

9. The method of claim 1, wherein, After performing a logical inference operation in the target entity network to generate a reasoning result associated with the target entity identifier, further including: Recording the history record and the inference result of each SWRL rule in the SWRL inference rule set being triggered; Based on the history record and feedback information of the user for the inference result, calculating the performance index of each SWRL inference rule; In the case where the performance index is lower than the set index threshold, generating a rule optimization prompt or a rule disablement prompt.

10. The method of claim 5, wherein, After obtaining the knowledge graph instance, further including: Storing the knowledge graph instance in the form of a resource description framework (RDF) file.

11. The method of claim 5, wherein, The business inference request is a loan business request; the ontology class includes at least one of a loan applicant, a loan application, a mortgage, a guarantor, and a credit score; and the class relationship includes at least one of an application loan relationship between the loan applicant and the loan application, and a mortgage relationship between the loan applicant and the mortgage.

12. The method of claim 10, wherein, The target entity identifier includes a loan applicant identifier or a loan business identifier.

13. A service inference apparatus characterized by comprising: Including: An entity identifier acquisition module for acquiring a target entity identifier included in a business inference request initiated by a user end in response to the business inference request; An entity network determination module for retrieving a target entity network matching the target entity identifier from a knowledge graph instance according to the target entity identifier; the knowledge graph instance is constructed based on a pre-constructed knowledge graph ontology, and is constructed by mapping and instantiating business data; the knowledge graph instance includes entities, entity attributes, and entity relationships; A logical inference module for loading a predefined semantic web rule language (SWRL) inference rule set, and performing a logical inference operation in the target entity network based on the SWRL inference rule set using an inference engine to generate a reasoning result associated with the target entity identifier; A knowledge graph optimization module for optimizing and updating the knowledge graph instance based on new entity attributes and new entity relationships included in the reasoning result; A reasoning result feedback module is configured to feed back the reasoning result to the user terminal through an application programming interface.

14. A business reasoning electronic device, comprising: The computer program is stored in the memory and executable in the processor, and the processor implements the business reasoning method according to any one of claims 1-12 when executing the computer program.

15. A computer readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the business reasoning method according to any one of claims 1-12.

16. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the business reasoning method according to any one of claims 1-12. The computer program is executed by the processor to implement the business reasoning method according to any one of claims 1-12.

Images