Method and device for updating map information and processor

Abstract

The invention discloses a map information updating method and device and a processor. The method comprises the following steps: in response to a detected reasoning instruction, obtaining a reasoning rule and a graph identifier from the reasoning instruction; calling atlas information corresponding to the atlas identifier from an atlas database, and detecting an initial association relationship between entity information in the atlas information to obtain a detection result; in response to the detection result that the initial association relationship does not meet the inference rule, constructing a target association relationship meeting the inference rule based on the inference rule and the entity information; and on the basis of the target association relationship, updating the map information to obtain updated target map information. According to the invention, the technical problem of low updating efficiency of the map information is solved.

Description

Technical Field

[0001] The present invention relates to the field of knowledge graph technology, and in particular to a method, device and processor for updating graph information. Background Art

[0002] At present, knowledge graph (graph information) as a new generation of knowledge representation model, by building a relationship network between entity information (entities), the complex relationship between entity information is made explicit, and the attribute reasoning of entity information is realized. It is widely used in artificial intelligence fields such as cognitive computing, semantic search, and intelligent decision-making. Therefore, it is very important to quickly and efficiently determine the relationship between entity information and develop a corresponding knowledge graph.

[0003] In related technologies, the association between entity information can be inferred based on inference rules to construct a knowledge graph. However, the current inference rules need to rely on clear logical rules and are complex to express. For example, professionals are required to manually write the inference code corresponding to the inference rules. This requires users to have a certain technical background and professional knowledge to define. The complexity of the inference rules and the knowledge requirements in specific fields limit the popularity of the inference rules and their use by non-professionals. In addition, in the inference stage of association relationships, since professionals often have complex descriptions of entity information and limited ability to express association relationship reasoning, in most cases, rules are manually written to perform association relationship reasoning, which usually requires the knowledge support of professionals. Not only can it not support more complex reasoning, but the correctness and consistency of the reasoning results are difficult to guarantee, and manual rule writing is time-consuming and labor-intensive. Therefore, there is still a technical problem of low efficiency in updating graph information.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] Embodiments of the present invention provide a method, device, and processor for updating graph information, so as to at least solve the technical problem of low efficiency in updating graph information.

[0006] According to one aspect of an embodiment of the present invention, a method for updating graph information is provided, comprising: in response to detecting an inference instruction, obtaining an inference rule and a graph identifier from the inference instruction; calling graph information corresponding to the graph identifier from a graph database, and detecting the initial association relationship between entity information in the graph information to obtain a detection result; in response to the detection result that the initial association relationship does not satisfy the inference rule, constructing a target association relationship that satisfies the inference rule based on the inference rule and the entity information; and updating the graph information based on the target association relationship to obtain updated target graph information.

[0007] Optionally, the method for updating graph information is applied to a graph information updating system, the system comprising an interaction layer and a reasoning layer, wherein in response to detecting a reasoning instruction, obtaining reasoning rules and graph identification from the reasoning instruction, including: detecting the reasoning instruction using the interaction layer; transmitting the reasoning instruction from the interaction layer to the reasoning layer in response to detecting the reasoning instruction; and obtaining the reasoning rules and graph identification from the reasoning instruction using the reasoning layer.

[0008] Optionally, calling graph information corresponding to a graph identifier from a graph database includes: using an inference layer to convert the acquired inference rules and graph identifiers to obtain target statements corresponding to the inference rules and graph identifiers, wherein the target statement is an executable statement of the graph database; and executing the target statement in the graph database to call the graph information indicated by the target statement.

[0009] Optionally, the inference rule includes an operation type, an entity relationship mapping and a derivation condition, and the operation type includes a query operation type, wherein the method further includes: in response to the operation type being a query operation type, converting the inference rule and the graph identifier into a target statement, wherein the target statement is an executable statement of the graph database; executing the target statement in the graph database, and screening the graph information in the graph information set in the graph database according to the entity relationship mapping and the derivation condition to obtain the graph information that satisfies the target statement.

[0010] Optionally, the inference rule includes an operation type and an entity relationship mapping, and the operation type includes an inference operation type, wherein the initial association relationship between the entity information in the graph information is detected to obtain a detection result, including: in response to the operation type being an inference operation type, the initial association relationship is detected based on the entity relationship mapping to obtain a detection result.

[0011] Optionally, the inference rule includes a derivation condition and a result relationship mapping, wherein, in response to the detection result that the initial association relationship does not satisfy the inference rule, a target association relationship that satisfies the inference rule is constructed based on the inference rule and the entity information, including: in response to the detection result that the initial association relationship does not satisfy the entity relationship mapping, based on the inference rule and the entity information, the target association relationship is constructed; the method also includes: filling the target association relationship into the result relationship mapping.

[0012] Optionally, the system includes a data layer, wherein, after updating the graph information based on the target association relationship to obtain updated target graph information, the method further includes: using the data layer to store the updated target graph information in a graph database, and deleting the graph information before the update corresponding to the target graph information in the graph database.

[0013] According to another aspect of an embodiment of the present invention, a device for updating graph information is also provided, including: an acquisition unit, for acquiring an inference rule and a graph identifier from an inference instruction in response to detecting an inference instruction; a detection unit, for calling the graph information corresponding to the graph identifier from a graph database, and detecting the initial association relationship between entity information in the graph information to obtain a detection result; a construction unit, for constructing a target association relationship that satisfies the inference rule based on the inference rule and entity information in response to the detection result that the initial association relationship does not satisfy the inference rule; and an update unit, for updating the graph information based on the target association relationship to obtain an updated target graph information.

[0014] According to another aspect of an embodiment of the present invention, a processor is further provided, which can be used to run a program, wherein the program executes any of the above-mentioned methods for updating graph information when it is run.

[0015] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing any one of the above-mentioned methods for updating graph information.

[0016] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any one of the above-mentioned methods for updating graph information.

[0017] According to another aspect of an embodiment of the present invention, a computer program product is provided, comprising computer instructions, which, when executed by a processor, implement any one of the above-mentioned methods for updating graph information.

[0018] In an embodiment of the present invention, if the graph information needs to be updated, when an inference instruction is detected, the inference rule and the graph identifier can be obtained from the inference instruction. The graph information corresponding to the graph identifier can be called from the graph database, and the initial association relationship between the entity information in the graph information can be detected to obtain the detection result. When the detection result is that the initial association relationship does not meet the inference rule, a target association relationship that meets the inference rule can be constructed based on the inference rule and the entity information. The graph information can be updated based on the target association relationship to obtain the updated target graph information. In this embodiment, the user can trigger the inference instruction according to his own needs, and obtain the inference rule and the graph identifier from the inference instruction without directly writing complex inference code. The graph information corresponding to the graph identifier can be called from the graph database, and the initial association relationship between the entity information in the graph information can be automatically detected. According to the inference rule and the entity information, the graph information can be automatically updated without manual intervention, thereby solving the technical problem of low efficiency in updating the graph information and achieving the technical effect of improving the efficiency of updating the graph information. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0020] Figure 1 is a flow chart of a method for updating graph information according to an embodiment of the present invention;

[0021] Figure 2 is a schematic diagram of a logical architecture of a rule-based knowledge graph relationship reasoning method according to an embodiment of the present invention;

[0022] Figure 3 is a schematic diagram of a reasoning module according to an embodiment of the present invention;

[0023] Figure 4 is a schematic diagram of a regular format design according to an embodiment of the present invention;

[0024] Figure 5 is a schematic diagram of a reasoning module architecture according to an embodiment of the present invention;

[0025] Figure 6 is a schematic diagram of rule conversion according to an embodiment of the present invention;

[0026] Figure 7 is a flow chart of a reasoning method according to an embodiment of the present invention;

[0027] Figure 8 is a schematic diagram of an example according to an embodiment of the present invention;

[0028] Fig. 9 is a structural schematic diagram of a device for updating graph information according to an embodiment of the present invention;

[0029] Fig.10 is a schematic diagram of an electronic device for a method for updating graph information according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] Example 1

[0033] According to an embodiment of the present invention, an embodiment of a method for updating graph information is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0034] Figure 1 is a flow chart of a method for updating graph information according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0035] Step S102, in response to detecting an inference instruction, obtaining an inference rule and a graph identifier from the inference instruction.

[0036] In the technical solution provided in the above step S102 of the embodiment of the present invention, the reasoning instruction can be triggered autonomously by the user according to his own needs. The reasoning rule can be a logical expression that derives new knowledge or verifies potential relationships based on existing knowledge, and can be a rule issued by the user. The graph identifier is a unique identifier of the knowledge graph, which can be used to distinguish and locate different knowledge graphs. The knowledge graph is a structured knowledge representation method that can describe entity information (such as people, places, events, concepts, etc.) and the relationship between them in the form of a graph. In the knowledge graph, entity information can usually be represented as a node, and the association relationship between entity information can be represented as an edge connecting the nodes. The structure of the above knowledge graph can intuitively show the connection and hierarchy between entity information. The entity information and association relationships in the knowledge graph can come from various data sources, including text, databases, web pages, etc.

[0037] In this embodiment, if an inference instruction is detected, the inference rule and graph identifier can be obtained from the inference instruction to determine which specific knowledge graph the inference operation will be performed on. The relevant data can be quickly located and the inference operation can be performed in a targeted manner, thereby greatly improving the inference efficiency.

[0038] Optionally, the inference rules can be predefined by the user according to business requirements and domain knowledge, and can be expressed using an easy-to-understand, high-level rule template language. This is for illustrative purposes only and is not specifically limited here.

[0039] For example, users can input or select rule template languages ​​to trigger inference instructions according to their own needs in the interactive interface, without the need for inference rules in the form of complex codes. For example, in the knowledge graph of entity information such as aircraft, airports, and routes, users can define the following inference rules: infer from aircraft to airport by station relationship; where aircraft has type = commercial aircraft; and airport has region = China; and count (landing events) > 5. This inference rule means that from the aircraft entity to the airport entity, if there is a station relationship (by station relationship), the following conditions must be met: the aircraft entity must be a commercial aircraft (aircraft has type = commercial aircraft), the airport entity is located in China (airport has region = China), and the number of landing events of the aircraft at the airport exceeds 5 times (count (landing events) > 5).

[0040] Step S104, calling the graph information corresponding to the graph identifier from the graph database, and detecting the initial association relationship between the entity information in the graph information to obtain the detection result.

[0041] In the technical solution provided in the above step S104 of the embodiment of the present invention, according to the inference instruction, after obtaining the inference rule and the graph identifier from the inference instruction, the graph information corresponding to the graph identifier can be called from the graph database, and the initial association relationship between the entity information in the graph information can be detected to obtain the detection result. The graph database can be a Janus Graph (JanusGraph, referred to as JG) database. JanusGraph is a high-performance, scalable graph database system that can handle large-scale graph data structures and support transaction processing, real-time query and complex relationship analysis. The graph information can be a knowledge graph.

[0042] In this embodiment, the graph identifier can point to a specific knowledge graph stored in the graph database. Through the graph identifier, the correct graph information can be located, and entity information, association relationship and other information can be read from the knowledge graph. After calling the graph information, the initial association relationship between the entity information can be detected, that is, whether the existing relationship between entities in the knowledge graph satisfies the inference rules to obtain the detection result.

[0043] For example, if the inference rule is to infer the stationing relationship between an aircraft and an airport, it can be detected whether the aircraft has a landing record at the airport and whether the landing frequency reaches the threshold defined by the rule. The above detection can be achieved by executing a specific Gramlin query.

[0044] Optionally, Greml in is a feature-rich graph traversal language developed by the graph computing framework TinkerPop, which can be used to explore, analyze, and manipulate data in graph databases. Greml in supports the execution of complex graph traversal logic on distributed graph databases, can process large-scale graph structured data, and allows users to define data query and update operations in a declarative or procedural manner. Executing Greml in statements in JanusGraph can complete the work of reasoning about association relationships in the knowledge base and filter out entity pairs that meet the inference rule conditions.

[0045] Optionally, the initial association relationship is the basis for building a knowledge graph, which can be used to connect different entity information to form a graph structure representing domain knowledge. By detecting the initial association relationship between entity information, it can ensure that the inference rules are performed on relevant and valid graph information, avoiding wasting computing resources on irrelevant entity pairs and improving the accuracy of inference.

[0046] Step S106, in response to the detection result that the initial association relationship does not satisfy the inference rule, a target association relationship that satisfies the inference rule is constructed based on the inference rule and the entity information.

[0047] In the technical solution provided in the above step S106 of the embodiment of the present invention, after the initial association relationship between the entity information in the graph information is detected and the detection result is obtained, if the detection result is that the initial association relationship does not satisfy the inference rule, a target association relationship that satisfies the inference rule can be constructed based on the inference rule and the entity information.

[0048] In this embodiment, when the detection result shows that the relationship between entities does not meet the direct premise of the current reasoning rule, the entity information can be further analyzed, and a new target association relationship that meets the rule can be constructed based on the deduction logic of the reasoning rule, so that the hidden potential relationship between entities can be quickly discovered. For example, if the reasoning rule is to infer the potential stationing relationship between the aircraft and the airport, and the detection result shows that the number of landings of the aircraft at the airport does not reach the threshold, the flight mode, stay time, and other relationships with the airport of the aircraft can be analyzed to construct a more comprehensive basis for judging the stationing relationship.

[0049] Optionally, by constructing target associations that satisfy inference rules, the knowledge graph can more comprehensively reflect the potential associations between entities, improving the depth and breadth of the knowledge graph. The newly constructed target associations can provide a richer and more accurate data foundation for intelligent analysis and decision-making, enabling better decisions to be made in complex scenarios.

[0050] Step S108, based on the target association relationship, the graph information is updated to obtain updated target graph information.

[0051] In the technical solution provided in the above step S108 of the embodiment of the present invention, the detection result is that the initial association relationship does not satisfy the inference rule. After constructing the target association relationship that satisfies the inference rule based on the inference rule and the entity information, the graph information can be updated based on the target association relationship to obtain the updated target graph information.

[0052] In this embodiment, if the target association relationship does not exist in the knowledge graph, the association relationship can be added and the attribute information of the entity can be updated accordingly. If the target association relationship already exists, but the attributes or weights need to be adjusted, the attribute information of the association relationship can be modified. In the case of inaccurate association relationships, the erroneous association relationship information can be deleted or modified to maintain the accuracy of the knowledge graph.

[0053] Optionally, based on the target relationship, the graph information can also be updated through incremental updates or full updates to integrate the target relationship into the existing graph information. Incremental updates can only add or modify entity and relationship information related to the new target relationship, while keeping the rest of the knowledge graph unchanged. Full updates rebuild the entire knowledge graph to ensure that the integration of new relationships does not affect the global consistency of the knowledge graph.

[0054] It should be noted that the above-mentioned updating method of the atlas information is only an example and is not specifically limited here. As long as it can be used to update the atlas information, it is within the protection scope of the embodiments of the present invention.

[0055] In the technical solution provided by the above steps S102 to S108 of the embodiment of the present invention, if the graph information needs to be updated, when the inference instruction is detected, the inference rule and the graph identifier can be obtained from the inference instruction. The graph information corresponding to the graph identifier can be called from the graph database, and the initial association relationship between the entity information in the graph information can be detected to obtain the detection result. When the detection result is that the initial association relationship does not meet the inference rule, a target association relationship that meets the inference rule can be constructed based on the inference rule and the entity information. The graph information can be updated based on the target association relationship to obtain the updated target graph information. In this embodiment, the user can trigger the inference instruction according to his own needs, and obtain the inference rule and the graph identifier from the inference instruction without directly writing complex inference code. The graph information corresponding to the graph identifier can be called from the graph database, and the initial association relationship between the entity information in the graph information can be automatically detected. According to the inference rule and the entity information, the graph information can be automatically updated without manual intervention, thereby solving the technical problem of low efficiency in updating the graph information and achieving the technical effect of improving the efficiency of updating the graph information.

[0056] The embodiment of the present invention is described in detail below in combination with the above steps.

[0057] As an optional embodiment, the graph information updating method is applied to a graph information updating system, the system includes an interaction layer and a reasoning layer, wherein step S102, obtaining inference rules and graph identification from inference instructions, includes: detecting the inference instructions using the interaction layer; in response to detecting the inference instructions, transmitting the inference instructions from the interaction layer to the reasoning layer; and obtaining the inference rules and graph identification from the inference instructions using the reasoning layer.

[0058] In this embodiment, the updating method of the graph information can be applied to the updating system of the graph information, and the system can include an interaction layer and a reasoning layer. If the reasoning rules and graph identification are to be obtained from the reasoning instructions, the interaction layer can be used to detect the reasoning instructions. When the reasoning instructions are detected, the reasoning instructions can be transmitted from the interaction layer to the reasoning layer. The reasoning layer can be used to obtain the reasoning rules and graph identification from the reasoning instructions. The interaction layer can complete the interaction function between the system and the user, and meet the user's needs of inputting rules according to the ontology model, viewing the rules in the rule library, etc. The reasoning layer can complete the entire process services such as rule conversion, execution, and return of reasoning results.

[0059] Optionally, the system can receive reasoning instructions input by the user at the interaction layer, encapsulate the instructions and transmit them to the reasoning layer to ensure the accuracy and security of the reasoning instructions, so as to protect the graph information from being changed by unauthorized users. After receiving the reasoning instructions, the reasoning layer can perform in-depth analysis to extract the reasoning rules and graph identifiers from them. Since the reasoning rules contain the logic and conditions for reasoning, and the graph identifiers indicate the specific graph information, the system can accurately locate the graph information that needs to be reasoned and updated.

[0060] As an optional embodiment, step S104, calling the graph information corresponding to the graph identifier from the graph database, includes: using the inference layer to convert the acquired inference rules and graph identifier to obtain the target statement corresponding to the inference rules and graph identifier, wherein the target statement is an executable statement of the graph database; executing the target statement in the graph database to call the graph information indicated by the target statement.

[0061] In this embodiment, if the graph information corresponding to the graph identifier is to be called from the graph database, the inference layer can be used to convert the acquired inference rules and graph identifiers to obtain the target statement corresponding to the inference rules and graph identifiers. For example, if the graph database uses JanusGraph and supports Greml in query language, the inference layer can convert the inference rules into corresponding Greml in statements to adapt to the query mechanism of the graph database. The target statement can be executed in the graph database to call the graph information indicated by the target statement. Among them, the target statement can be an executable statement of the graph database, such as a Greml in statement.

[0062] Optionally, during the conversion process, the inference layer can map the key elements of the inference rules, such as entities, relationships, attributes, and inference conditions, into the database language to generate specific target statements. For example, if the inference rule involves finding entities with specific attributes and checking whether the relationship between them satisfies a certain condition, the generated Greml in statement will contain the corresponding node and edge query instructions, as well as attribute filtering and conditional judgment logic.

[0063] Optionally, the target statement is executed in the graph database, and the execution process involves the query engine of the database, which can parse and execute the target statement and call the graph information that meets the conditions from the database. For Greml in statements, the execution process can include graph traversal, path finding, attribute matching and other operations to locate and obtain entity and relationship data related to the inference rule.

[0064] Optionally, the execution result of the target statement is the graph information corresponding to the graph identifier, which can be further processed by the reasoning layer for subsequent reasoning judgment and update operations. The acquired graph information may include the attributes of entities, the association relationship between entities, and events or data points related to reasoning, etc., which are only for example and are not specifically limited here.

[0065] As an optional embodiment, the inference rule includes an operation type, an entity relationship mapping and a derivation condition, and the operation type includes a query operation type, wherein the method further includes: in response to the operation type being a query operation type, converting the inference rule and the graph identifier into a target statement, wherein the target statement is an executable statement of the graph database; executing the target statement in the graph database, and screening the graph information in the graph information set in the graph database according to the entity relationship mapping and the derivation condition to obtain the graph information that satisfies the target statement.

[0066] In this embodiment, the inference rule may include an operation type, an entity relationship mapping, and a derivation condition, and the operation type may include a query operation type. If the operation type is a query operation type, the inference rule and the graph identifier may be converted into a target statement. The target statement may be executed in the graph database, and the graph information in the graph information set in the graph database may be screened according to the entity relationship mapping and the derivation condition to obtain the graph information that satisfies the target statement. Among them, the target statement may be an executable statement of the graph database.

[0067] Optionally, if the operation type is a query operation type, the inference layer can convert the inference rule into a query statement executable by the graph database, namely, a target statement. In the process of executing the target statement in the graph database, the graph database can filter the stored graph information set according to the entity relationship mapping and derivation conditions to retrieve the graph information of the target statement that meets the query conditions. For example, if the inference rule requires finding all aircraft that have stopped at a specific airport more than three times, the generated Greml in statement will traverse the graph database and perform matching and filtering based on the landing event relationship between the aircraft entity and the airport entity and its frequency information.

[0068] As an optional embodiment, the inference rules include operation types and entity relationship mappings, and the operation types include reasoning operation types. In step S104, the initial association relationship between entity information in the graph information is detected to obtain a detection result, including: in response to the operation type being a reasoning operation type, the initial association relationship is detected based on the entity relationship mapping to obtain a detection result.

[0069] In this embodiment, the inference rule may include an operation type and an entity relationship mapping, and the operation type may include an inference operation type. If the initial association relationship between entity information in the graph information is to be detected and a detection result is obtained, the initial association relationship may be detected based on the entity relationship mapping when the operation type is an inference operation type to obtain a detection result.

[0070] Optionally, the operation type can be used to distinguish the specific actions of the current rule, which can be divided into query actions and inference actions. Query actions can be distinguished using the query keyword, and inference actions can be distinguished using the infer keyword. Entity-relationship mapping can be performed by keywords such as from A to B by Event. For example, in a knowledge graph based on airplanes and airports, "from airplane A to airport B by landing event" means the event that airplane A lands at airport B, thereby establishing an entity relationship of landing between airplane A and airport B.

[0071] As an optional embodiment, the inference rule includes a derivation condition and a result relationship mapping, wherein, in step S106, in response to the detection result that the initial association relationship does not satisfy the inference rule, based on the inference rule and the entity information, a target association relationship that satisfies the inference rule is constructed, including: in response to the detection result that the initial association relationship does not satisfy the entity relationship mapping, based on the inference rule and the entity information, a target association relationship is constructed; the method also includes: filling the target association relationship into the result relationship mapping.

[0072] In this embodiment, the inference rule may include a derivation condition and a result relationship mapping. If the detection result is that the initial association relationship does not satisfy the inference rule, a target association relationship that satisfies the inference rule is constructed based on the inference rule and the entity information. If the detection result is that the initial association relationship does not satisfy the entity relationship mapping, the target association relationship may be constructed based on the inference rule and the entity information. The target association relationship may be filled into the result relationship mapping.

[0073] Optionally, the derivation conditions can be used to further derive the results of the previous derivation, because for some conditional derivations, the use of database query statements is not fully competent. For example, the results can be grouped and attribute range filtered, as well as some customized judgment operations. The results can be processed in a program coding manner, and the data can be extracted from the database for further processing to obtain the complete results. The result relationship mapping can be used to specify the meaning of the relationship between the entities that meet the above conditional rules in the final generated result set. Users can annotate this type of association with a label, and finally store the association back to the graph database.

[0074] As an optional embodiment, the system includes a data layer, wherein, after updating the graph information based on the target association relationship to obtain the updated target graph information, the method also includes: using the data layer to store the updated target graph information in a graph database, and deleting the graph information before the update corresponding to the target graph information in the graph database.

[0075] In this embodiment, in addition to the interaction layer and the reasoning layer, the system may also include a data layer. The data layer may be used to complete the storage and acquisition of rules and the storage and acquisition of knowledge. After updating the graph information based on the target association relationship and obtaining the updated target graph information, the data layer may be used to store the updated target graph information in the graph database and delete the graph information before the update corresponding to the target graph information in the graph database.

[0076] Optionally, after the inference layer completes the inference operation and updates the graph information based on the target association relationship, the updated target graph information can be passed to the data layer. After the data layer receives the updated target graph information, it can store it in the graph database. The storage process can include write operations to the graph database to ensure that information such as new relationships, entity attributes, or events are correctly added to the knowledge graph to reflect the results of the inference layer inference. During the storage process, the data layer can ensure the consistency and security of the data based on the characteristics of the graph database, such as transaction processing, concurrency control, etc.

[0077] Optionally, in order to maintain the clarity of the knowledge graph and avoid data redundancy, the data layer can delete the corresponding graph information before the update in the graph database while storing the updated target graph information, thereby more accurately reflecting the latest reasoning results and avoiding the existence of outdated or conflicting data.

[0078] In an embodiment of the present invention, if the graph information needs to be updated, when an inference instruction is detected, the inference rule and the graph identifier can be obtained from the inference instruction. The graph information corresponding to the graph identifier can be called from the graph database, and the initial association relationship between the entity information in the graph information can be detected to obtain the detection result. When the detection result is that the initial association relationship does not meet the inference rule, a target association relationship that meets the inference rule can be constructed based on the inference rule and the entity information. The graph information can be updated based on the target association relationship to obtain the updated target graph information. In this embodiment, the user can trigger the inference instruction according to his own needs, and obtain the inference rule and the graph identifier from the inference instruction without directly writing complex inference code. The graph information corresponding to the graph identifier can be called from the graph database, and the initial association relationship between the entity information in the graph information can be automatically detected. According to the inference rule and the entity information, the graph information can be automatically updated without manual intervention, thereby solving the technical problem of low efficiency in updating the graph information and achieving the technical effect of improving the efficiency of updating the graph information.

[0079] Example 2

[0080] Another optional specific implementation is described in detail below.

[0081] At present, knowledge graphs are playing an increasingly important role as a powerful knowledge representation and management tool. Relational reasoning in knowledge graphs aims to discover new relationships or verify potential relationships from existing knowledge. Rule-based reasoning methods are an important technology in relational reasoning in knowledge graphs. In related technologies, reasoning usually relies on clear logical rules, which are usually manually defined by domain experts. The rules are encoded into a specific query language and executed in the knowledge graph to discover new entity relationships or verify potential connections. Rule-based reasoning requires users to have a certain technical background and expertise to define rules. The complexity of the rules and the knowledge requirements in specific fields limit the popularity of rule reasoning and the use of non-professionals. Therefore, there is still a technical problem of low efficiency in updating knowledge graphs.

[0082] The present invention proposes a rule-based knowledge graph relationship reasoning method and system. A rule system that is more user-friendly can be provided so that non-expert users can also define new reasoning rules according to business needs, and apply the above reasoning rules to the knowledge graph to realize rapid discovery of potential relationships between entities, thereby achieving efficient reasoning and rapid discovery of hidden potential relationships between entities. This solves the technical problem of low update efficiency of graph information and achieves the technical effect of improving the update efficiency of graph information.

[0083] The method is further described below.

[0084] Figure 2 is a schematic diagram of the logical architecture of a rule-based knowledge graph relationship reasoning method according to an embodiment of the present invention, such as Figure 2 As shown, it may include: an interaction layer 201 , a reasoning layer 202 and a data layer 203 .

[0085] The interaction layer 201 can be used to complete the interaction function between the file system module (file system a, file system b) and the user, and meet the user's needs of inputting rules according to the ontology model and viewing the rules in the rule base.

[0086] The reasoning layer 202 can be used to complete the entire process services such as rule conversion, execution, and return of reasoning results, and can be divided into rule conversion analysis, conflict resolution, and rule execution. Rule conversion analysis can convert the rules entered by the user into Greml in statements. Conflict resolution can solve the problem of conflicts when multiple rules are executed simultaneously. Rule execution can execute Greml in statements in JanusGraph (such as JG_A, JG_B) to complete the work of reasoning new knowledge in the knowledge base.

[0087] The data layer 203 can be used to store and retrieve rules and store and retrieve knowledge.

[0088] Figure 3 is a schematic diagram of a reasoning module according to an embodiment of the present invention, such as Figure 3 As shown, it may include: an entry module 301, an execution management module 302, a conversion module 303 and a storage management module 304. The entry module 301 that interacts with the user may receive rules and rule management commands input by the user. The formatted rules are transmitted to the conversion module 303, and the conversion module 303 may transmit the Greml in-rule pairs generated after the rule conversion to the storage management module 304 to prepare for the storage of the rule storage file. The storage module 304 may accept the add and delete operation commands forwarded from the entry module 301 to store the Gremlin-rule pairs. The execution management module 302 may perform reasoning operations on the Greml in-rule pairs received from the upstream functional module, interact with JanusGraph through the Greml in command, and complete the reasoning.

[0089] In order to enable users to use rule derivation more efficiently and avoid using native graph database operation statements, the embodiment of the present invention designs a set of concise and easy-to-use rule formats for users to establish and derive rules. Figure 4 is a schematic diagram of a regular format design according to an embodiment of the present invention, such as Figure 4As shown, it includes: an operation type module 401, an entity relationship mapping module 402, a derivation condition module 403 and a result relationship mapping module 404.

[0090] The operation type module 401 can be used to distinguish the specific actions of the current rule, which are mainly divided into query actions and inference actions. The query actions can be distinguished using the query keyword, and the inference actions can be distinguished using the infer keyword.

[0091] The entity relationship mapping module 402 can have different relationships for different entities according to their specific social attributes or events. In the rules, entities and relationships can be specified using keywords such as from A to B byEvent.

[0092] The derivation condition module 403 can be used to further derive the results of the previous derivation, because for some conditional derivations, the use of database query statements is not fully competent. For example, the results can be grouped and attribute range filtered, as well as some customized judgment operations. The results can be processed by program coding, and the data can be extracted from the database for further processing to finally obtain the complete result. The syntax of the derivation condition is similar to the way of calling a function. The derivation condition can be a custom implementation method, and different derivation conditions have different numbers of parameter inputs. The grammar rules are as follows:

[0093] [attr operator arg;attr function(arg1,arg2,…)]

[0094] Among them, attr can be used to represent attributes, which are characteristics or features of entities in the rule, such as the entity's name, location, time, event frequency, etc. operator can be used to represent operators, which are used to compare or perform logical operations on attributes. arg can be used to represent parameters or values, which are used to compare or operate with attributes. function can be used to represent functions, which are used to perform more complex operations or calculations on one or more attributes. (arg1, arg2, ...) can be used to represent a function's parameter list.

[0095] The result relationship mapping module 404 can be used to specify the meaning of the relationship between entities that meet these conditional rules in the final generated result set. The user can mark this type of relationship with a label and finally store the above relationship back into the graph database.

[0096] Figure 5 is a schematic diagram of a reasoning module architecture according to an embodiment of the present invention, such as Figure 5 As shown, it includes: a user interaction layer 501, a rule reasoning layer 502, and a data storage layer 503.

[0097] The user interaction layer 501 can complete the interaction function between the file system module and the user, and meet the user's needs of inputting rules according to the ontology model and viewing the rules in the rule library through the rule input view and the rule management view.

[0098] The rule reasoning layer 502 can complete the entire process services such as rule conversion, execution, and return of reasoning results, etc. It can include functions such as ontology model reading, rule management, and rule conversion.

[0099] In this embodiment, in the Linux operating system, the construction of the reasoning module architecture can be achieved through the Java development kit. Jena can be used to read the ontology model, and the created ontology model file is parsed through the file input / output application program interface to complete the reading work. Jena is a Java toolbox for developing semantic network applications based on the Resource Description Framework (RDF) and the Web Ontology Language (OWL), which can be used to manage ontology models stored in relational databases or text files. TinkerPop can run on a variety of graph databases, such as JanusGraph. Rule management can add, delete, modify, and query the management of user data based on established rule definitions. Rule conversion can include parsers and generators.

[0100] The data storage layer 503 can store file systems and graph databases.

[0101] Figure 6 is a schematic diagram of a rule conversion according to an embodiment of the present invention, such as Figure 6 As shown, it includes: a parser 601 and a generator 602.

[0102] Parser 601 can be used to process user input rules. The upper-level high-level statements will undergo grammatical and semantic analysis through the processing of the parsing layer, and the rule syntax parsing will be performed on the rule logic that the user wants to execute, and finally saved in the parsed data structure to facilitate the execution of generator 602.

[0103] Generator 602 can utilize semantic information generated by parser 601 in generator 602, combined with grammatical rules of Greml in to generate corresponding executable Greml in statements, and after analysis and optimization, avoid some unnecessary information, and generate a set of executable Greml in statements.

[0104] Figure 7 is a flow chart of a reasoning method according to an embodiment of the present invention. Figure 7As shown, the following steps are included:

[0105] Step S701: whether the rule is valid.

[0106] In this embodiment, the user can judge whether the rule is valid based on the inference service and the natural rule. If it is valid, step S702 can be executed; otherwise, an error is returned and the service ends.

[0107] Step S702, formatting.

[0108] In this embodiment, when the rule is valid, the formatting operation can be performed.

[0109] Step S703: extract keywords.

[0110] In this embodiment, after the formatting operation is performed, keyword extraction may be performed according to the formatting rules.

[0111] Step S704, generate Greml in.

[0112] In this embodiment, after keyword extraction, Gremlins may be generated.

[0113] Step S705, reading and writing files.

[0114] In this embodiment, files can be read and written according to Gremlin rules.

[0115] Step S706: whether to start reasoning.

[0116] In this embodiment, when reading and writing files and storing status receipts, it can be determined whether to start reasoning. If so, step S707 can be executed; otherwise, the service ends.

[0117] Step S707, read and write files.

[0118] In this embodiment, if reasoning can be started, file reading and writing are performed.

[0119] Step S708: perform priority determination.

[0120] In this embodiment, after reading and writing the file, a priority determination may be performed.

[0121] Step S709, perform JG reasoning operation.

[0122] In this embodiment, after the priority determination is performed, a JanusGraph reasoning operation can be performed to obtain a result receipt and an operation log.

[0123] In this embodiment, the rule business reasoning process can be divided into four parts based on the user usage process. The first part is the entry stage, that is, the user customizes the reasoning or query rules according to the established keywords and the existing entity business relationships in the knowledge graph, and the rules can be stored in the database system after parsing and verification. The second part is the conversion stage, which is to convert the rules entered by the user into a series of Greml in execution statements, and persistently map and store the Greml in execution statements. The third part is the access stage, that is, before executing the reasoning again, query the specified rule to obtain its corresponding reasoning statement. The fourth part is the reasoning stage. The reasoning engine module will execute the Greml in statements in sequence according to the rule reasoning statements to be executed in the queue and apply the execution results to the JanusGraph graph database, and finally complete the execution of the reasoning statement and the establishment of the new relationship.

[0124] Figure 8 is a schematic diagram of an example according to an embodiment of the present invention, such as Figure 8 As shown, in the knowledge graph based on aircraft entity 1, aircraft entity 2, ship entity 3, airport entity 4, airport entity 5, and island reef as core entities, it is inferred whether aircraft entity 1 and airport entity 5 are in a stopover or stationed relationship. Aircraft entity 1 and aircraft entity 2 contain corresponding aircraft trajectory information. Ship entity 3 contains corresponding ship trajectory information. Airport entity 4 and airport entity 5 contain corresponding airport locations. Island reef contains corresponding island reef location.

[0125] The stopover and stationing relationship describes the relationship between the interaction frequency between different aircraft entities and different airport entities (such as aircraft entity 1 and airport entity 5, aircraft entity 2 and airport entity 4). The events that can be captured between the two and exist in the knowledge graph are generally take-off events and landing events. Take-off events and landing events contain corresponding event times. A spatiotemporal relationship between the two can be inferred based on the occurrence characteristics of take-off events and landing events. If an aircraft entity (aircraft entity 1) frequently takes off and lands at the same airport entity (airport entity 5), it can be said that there is a permanent relationship between the two, and it can be considered that the aircraft entity (aircraft entity 1) is in a stationing relationship with the airport entity (airport entity 5). On the contrary, if the frequency of events between the two is low, it can be considered that the two are only in a stopover relationship. Then the rule template can judge the relationship between the two based on the frequency of events between the two. If an aircraft entity (aircraft entity 1) lands at an airport entity (airport entity 5) more than a certain number of times (for example, more than 3 times), it can be said that the aircraft entity (aircraft entity 1) is stationed at the airport entity (airport entity 5); otherwise, it is a stopover relationship.

[0126] Example 3

[0127] The embodiment of the present invention provides a device for updating graph information. It should be noted that the device for updating graph information in the embodiment of the present invention can be used to execute Figure 1 The method for updating the graph information provided by the embodiment of the present invention is described below. The device for updating the graph information provided by the embodiment of the present invention is described below.

[0128] Fig. 9 is a schematic diagram of a structure of a device for updating graph information according to an embodiment of the present invention. Fig. 9 As shown, the updating device 900 of the graph information may include: an acquisition unit 902, a detection unit 904, a construction unit 906 and an updating unit 908.

[0129] The acquisition unit 902 is used to acquire the inference rule and the graph identification from the inference instruction in response to detecting the inference instruction.

[0130] The detection unit 904 is used to call the graph information corresponding to the graph identifier from the graph database, and detect the initial association relationship between the entity information in the graph information to obtain the detection result.

[0131] The construction unit 906 is used to construct a target association relationship that satisfies the inference rule based on the inference rule and the entity information in response to the detection result that the initial association relationship does not satisfy the inference rule.

[0132] The updating unit 908 is used to update the graph information based on the target association relationship to obtain updated target graph information.

[0133] The graph information updating device provided by the embodiment of the present invention is used, through an acquisition unit 902, to obtain an inference rule and a graph identifier from the inference instruction in response to detecting an inference instruction; through a detection unit 904, to call the graph information corresponding to the graph identifier from the graph database, and to detect the initial association relationship between the entity information in the graph information to obtain a detection result; through a construction unit 906, to respond to the detection result that the initial association relationship does not satisfy the inference rule, based on the inference rule and the entity information, to construct a target association relationship that satisfies the inference rule; through an update unit 908, to update the graph information based on the target association relationship to obtain updated target graph information, thereby solving the technical problem of low updating efficiency of the graph information and achieving the technical effect of improving the updating efficiency of the graph information.

[0134] The above-mentioned atlas information updating device may also include a processor and a memory. The above-mentioned units are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0135] The processor includes a kernel, and the kernel retrieves the corresponding program unit from the memory. One or more kernels can be set to control the same device type of devices to be shut down to perform graceful shutdown by adjusting kernel parameters.

[0136] The above-mentioned memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0137] The processor contains a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the work efficiency of traders can be improved by adjusting the kernel parameters.

[0138] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0139] Example 4

[0140] According to an embodiment of the present invention, there is also provided a computer-readable storage medium on which a program is stored, and when the program is executed by a processor, a method for updating graph information is implemented.

[0141] Example 5

[0142] According to an embodiment of the present invention, a processor is also provided, and the processor is used to run a program, wherein the method for updating the graph information is executed when the program is running.

[0143] Example 6

[0144] Fig.10 is a schematic diagram of an electronic device for updating a method for graph information according to an embodiment of the present invention, such as Fig.10 As shown, an embodiment of the present invention further provides an electronic device, which includes a processor, a memory, and a program stored in the memory and executable on the processor, and a method for updating graph information is implemented when the processor executes the program.

[0145] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0146] Example 7

[0147] The invention also provides a computer program product which, when executed on a data processing device, is suitable for executing a program of the updating method for initializing graph information.

[0148] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0149] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0150] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0151] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0152] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0153] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0154] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transit media), such as modulated data signals and carrier waves.

[0155] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0156] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0157] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for updating graph information, characterized in that: include: In response to detecting the inference instruction, obtaining an inference rule and a graph identifier from the inference instruction; Retrieving the graph information corresponding to the graph identifier from the graph database, and detecting the initial association relationship between the entity information in the graph information to obtain a detection result; In response to the detection result that the initial association relationship does not satisfy the inference rule, constructing a target association relationship that satisfies the inference rule based on the inference rule and the entity information; Based on the target association relationship, the graph information is updated to obtain updated target graph information.

2. The method according to claim 1, characterized in that The method for updating graph information is applied to a graph information updating system, the system comprising an interaction layer and an inference layer, wherein, in response to detecting an inference instruction, an inference rule and a graph identifier are obtained from the inference instruction, including: Using the interaction layer to detect the reasoning instruction; In response to detecting the inference instruction, transmitting the inference instruction from the interaction layer to the inference layer; The inference layer is used to obtain the inference rule and the graph identifier from the inference instruction.

3. The method according to claim 2, characterized in that Retrieving the graph information corresponding to the graph identifier from the graph database includes: The acquired inference rule and the graph identifier are converted by using the inference layer to obtain a target statement corresponding to the inference rule and the graph identifier, wherein the target statement is an executable statement of the graph database; The target statement is executed in the graph database to call the graph information indicated by the target statement.

4. The method according to claim 2, characterized in that: The inference rule includes an operation type, an entity relationship mapping and a derivation condition, the operation type includes a query operation type, wherein the method further includes: In response to the operation type being the query operation type, converting the inference rule and the graph identifier into a target statement, wherein the target statement is an executable statement of the graph database; The target statement is executed in the graph database, and the graph information in the graph information set in the graph database is screened according to the entity relationship mapping and the derivation condition to obtain the graph information that satisfies the target statement.

5. The method according to claim 2, characterized in that: The inference rule includes an operation type and an entity relationship mapping, wherein the operation type includes an inference operation type, wherein the initial association relationship between the entity information in the graph information is detected to obtain a detection result, including: In response to the operation type being the reasoning operation type, the initial association relationship is detected based on the entity relationship mapping to obtain the detection result.

6. The method according to claim 5, characterized in that The inference rule includes a derivation condition and a result relationship mapping, wherein, in response to the detection result that the initial association relationship does not satisfy the inference rule, a target association relationship satisfying the inference rule is constructed based on the inference rule and the entity information, including: In response to the detection result that the initial association relationship does not satisfy the entity relationship mapping, constructing the target association relationship based on the inference rule and the entity information; The method further includes: filling the target association relationship into the result relationship mapping.

7. The method according to claim 2, characterized in that The system includes a data layer, wherein after the graph information is updated based on the target association relationship to obtain updated target graph information, the method further includes: The updated target graph information is stored in the graph database by using the data layer, and the graph information before the update corresponding to the target graph information in the graph database is deleted.

8. A device for updating graph information, characterized in that: include: an acquisition unit, configured to acquire an inference rule and a graph identifier from the inference instruction in response to detecting the inference instruction; A detection unit, used to call the graph information corresponding to the graph identifier from a graph database, and detect the initial association relationship between the entity information in the graph information to obtain a detection result; A construction unit, configured to construct, in response to the detection result that the initial association relationship does not satisfy the inference rule, a target association relationship satisfying the inference rule based on the inference rule and the entity information; An updating unit is used to update the graph information based on the target association relationship to obtain updated target graph information.

9. A processor, characterized in that: The processor is used to run a program, wherein the program, when run by the processor, executes the method for updating the graph information described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method for updating the graph information described in any one of claims 1 to 7.

Images