Knowledge graph construction method, tool, device and server

Abstract

This specification provides a method, tool, device and server for constructing a knowledge graph. Based on the knowledge graph construction method, the data structure type of the target source data to be processed can be determined first; then, according to the preset construction rules and the data structure type of the target source data, a target knowledge extraction unit matching the target source data can be constructed; further, the above-mentioned target knowledge extraction unit can be called to specifically process the target source data to obtain an entity relationship file that meets the requirements and contains multiple triple data groups; then, according to the above-mentioned entity relationship file, a target knowledge graph associated with the target source data is constructed. This can effectively simplify user-side operations and reduce the difficulty of constructing a knowledge graph, so that users can efficiently and accurately construct a knowledge graph that meets diverse business needs and has better effects.

Description

Technical Field

[0001] This specification belongs to the field of artificial intelligence technology, and in particular to methods, tools, devices and servers for constructing knowledge graphs. Background Art

[0002] Knowledge graph is an important branch of artificial intelligence technology and plays an important role in machine learning and cognition.

[0003] However, the existing knowledge graph construction methods have high technical barriers and are difficult to build for users who need to build knowledge graphs. In addition, based on the existing knowledge graph construction methods, when building knowledge graphs, there are often problems such as complex and cumbersome operations, low construction efficiency, and inability to meet the diverse business needs of users.

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

[0005] This specification provides a method, tool, device and server for constructing a knowledge graph to simplify user-side operations and reduce the difficulty of constructing a knowledge graph, so that users can efficiently and accurately construct a knowledge graph that meets diverse business needs and has better effects.

[0006] The embodiment of this specification provides a method for constructing a knowledge graph, including:

[0007] Get target source data;

[0008] Determine the data structure type of the target source data;

[0009] Constructing a target knowledge extraction unit matching the target source data according to a preset construction rule and a data structure type of the target source data;

[0010] Calling the target knowledge extraction unit to process the target source data to obtain an entity relationship file that meets the requirements; wherein the entity relationship file includes a plurality of triple data groups; the triple data group includes at least two data objects connected by a data relationship;

[0011] According to the entity relationship file, a target knowledge graph associated with the target source data is constructed.

[0012] In some embodiments, the data structure type of the target source data includes at least one of the following: structured data, unstructured data, and semi-structured data.

[0013] In some embodiments, constructing a target knowledge extraction unit matching the target source data according to a preset construction rule and a data structure type of the target source data includes:

[0014] According to a preset construction rule, a target source operator corresponding to the target source data is screened out from a plurality of preset data source operators; wherein the target source operator is used to connect the target source data to a target knowledge extraction unit;

[0015] Determine a matching target data processing structure according to the data structure type of the target source data; wherein the target data processing structure is used to process the target source data to obtain a plurality of triple data groups;

[0016] Determine and configure a target identification termination operator; wherein the target identification termination operator is used to extract a triple data group that meets the requirements from a plurality of triple data groups output by the target data processing structure to obtain a corresponding entity relationship file;

[0017] The target source operator, the target data processing structure and the target identification termination operator are combined to obtain a target knowledge extraction unit that matches the target data source.

[0018] In some embodiments, determining a matching target data processing structure according to the data structure type of the target source data includes:

[0019] When it is determined that the data structure type of the target source data is structured data, an initial processing operator is selected from a plurality of preset data processing operators;

[0020] The initial processing operator is configured accordingly to obtain a target processing operator; and the target processing operator is determined as a matching target data processing structure.

[0021] In some embodiments, the preset data processing operator includes at least one of the following: a SQL operator, a HIVE operator, and a SPARK operator.

[0022] In some embodiments, determining a matching target data processing structure according to the data structure type of the target source data includes:

[0023] When it is determined that the data structure type of the target source data is unstructured data or semi-structured data, a preset triple extraction model is determined as a matching target data processing structure.

[0024] In some embodiments, after determining the data structure type of the target source data, the method further includes:

[0025] According to the data structure type of the target source data, selecting a recommended knowledge extraction unit from a plurality of preset knowledge extraction units;

[0026] Displaying the recommended knowledge extraction unit to a user;

[0027] The recommended knowledge extraction unit selected by the user is determined as the target knowledge extraction unit.

[0028] In some embodiments, constructing a target knowledge graph associated with the target source data according to the entity relationship file includes:

[0029] Obtain a definition parameter file about the target knowledge graph; wherein the definition parameter file includes: definition parameters of data objects and / or definition parameters of data relationships;

[0030] According to the entity relationship file and the definition parameter file, a target knowledge graph associated with the target data source is generated by performing data mapping.

[0031] In some embodiments, the definition parameter file also includes index definition parameters;

[0032] Accordingly, in the process of constructing a target knowledge graph associated with the target source data according to the entity relationship file, the method further includes:

[0033] According to the index definition parameters, a target query index for the target knowledge graph is constructed using definition parameters of the data object and / or definition parameters of the data relationship.

[0034] In some embodiments, after constructing a target knowledge graph associated with the target source data according to the entity relationship file, the method further includes:

[0035] Receive a target query statement; wherein the target query statement carries at least a target identifier of a target knowledge graph;

[0036] According to the target identifier, searching the graph database to determine the target knowledge graph;

[0037] In response to the target query statement, a query operation is performed on the target knowledge graph to obtain a corresponding query result;

[0038] Feedback the query result.

[0039] In some embodiments, the target source data includes a flow record of a customer's transaction data; correspondingly, the query result includes a flow diagram of the target customer's transaction data.

[0040] The embodiment of this specification also provides a knowledge graph construction tool, which at least includes: a source data import interface, a first processing interface, and a second processing interface; wherein,

[0041] The source data import interface is used to support users to import target source data;

[0042] The first processing interface is used to support the user to set definition parameters of data objects and / or definition parameters of data relationships in the target knowledge graph to generate a definition parameter file for the target knowledge graph;

[0043] The second processing interface is used to support the user to determine and combine matching target source operators, target data processing structures, and identification termination operators according to preset construction rules to obtain target knowledge extraction units that match target source data;

[0044] The knowledge graph construction tool is also used to call the target knowledge extraction unit to process the target source data to obtain the corresponding entity relationship file; and based on the entity relationship file and the definition parameter file, generate a target knowledge graph associated with the target data source by performing data mapping.

[0045] The embodiment of this specification also provides a knowledge graph construction device, including:

[0046] An acquisition module is used to acquire target source data;

[0047] A determination module, used to determine the data structure type of the target source data;

[0048] A first construction module is used to construct a target knowledge extraction unit matching the target source data according to a preset construction rule and a data structure type of the target source data;

[0049] A calling module, used for calling the target knowledge extraction unit to process the target source data to obtain an entity relationship file that meets the requirements; wherein the entity relationship file contains a plurality of triple data groups; the triple data group includes at least two data objects connected by a data relationship;

[0050] The second construction module is used to construct a target knowledge graph associated with the target source data based on the entity relationship file.

[0051] An embodiment of the present specification also provides a server, including a processor and a memory for storing processor executable instructions, wherein the processor executes the steps of the knowledge graph construction method.

[0052] The embodiments of this specification also provide a computer storage medium on which computer instructions are stored, and the instructions, when executed, implement the steps of the method for constructing the knowledge graph.

[0053] This specification provides a method, tool, device and server for constructing a knowledge graph. Based on the method for constructing the knowledge graph, the data structure type of the target source data to be processed can be determined first; then, according to the preset construction rules and the data structure type of the target source data, a target knowledge extraction unit that matches the target source data and is more targeted can be constructed; further, the above-mentioned target knowledge extraction unit can be called to specifically process the target source data to obtain an entity relationship file that meets the requirements and contains multiple triple data groups; then, according to the above-mentioned entity relationship file, a target knowledge graph associated with the target source data can be constructed. This can effectively simplify user-side operations, reduce the difficulty of constructing the knowledge graph, and enable users to efficiently and accurately construct a knowledge graph that meets diverse business needs and has better effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of this specification, the drawings required for use in the embodiments will be briefly introduced below. The drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0055] Figure 1 It is a schematic diagram of an embodiment of the structural composition of a system using the knowledge graph construction method provided in the embodiments of this specification;

[0056] Figure 2 is a schematic diagram of a knowledge graph construction tool provided by an embodiment of this specification;

[0057] Figure 3 It is a flowchart of a method for constructing a knowledge graph provided by an embodiment of this specification;

[0058] Figure 4 It is a schematic diagram of the structure of a server provided by an embodiment of this specification;

[0059] Figure 5 It is a schematic diagram of the structure of a knowledge graph construction device provided by an embodiment of this specification;

[0060] Figure 6 It is a schematic diagram of an embodiment of the method for constructing a knowledge graph provided in an embodiment of this specification applied in a specific scenario example. DETAILED DESCRIPTION

[0061] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0062] The present specification provides a method for constructing a knowledge graph, which can be applied to a system including a server and a terminal device. Figure 1 The server and the terminal device may be connected via wired or wireless means to perform specific data interaction.

[0063] In this embodiment, the server may specifically include a background server applied to one side of the network platform and capable of realizing functions such as data transmission and data processing. Specifically, the server may be, for example, an electronic device having data calculation, storage and network interaction functions. Alternatively, the server may also be a software program running in the electronic device to provide support for data processing, storage and network interaction. In this embodiment, the number of servers included in the server is not specifically limited. The server may specifically be one server, or several servers, or a server cluster formed by several servers.

[0064] In this embodiment, the terminal device may specifically include a front-end electronic device applied to the user side and capable of realizing functions such as data collection and data transmission. Specifically, the terminal device may be, for example, a desktop computer, a tablet computer, a laptop computer, a smart phone, etc. Alternatively, the terminal device may also be a software application that can be run in the above electronic devices. For example, it may be an APP running on a smart phone.

[0065] In this embodiment, the server may be connected to a graph database of the network platform for maintaining and managing the graph database of the network platform. The graph database may store multiple knowledge graphs. The terminal device may be equipped with a knowledge graph construction tool.

[0066] In this embodiment, when the current user needs to process a batch of data (for example, the transaction data flow records of XX Bank's customers in 2020) to construct a target knowledge graph that meets business needs (for example, suitable for analyzing whether customers have risks of illegal transactions), a startup instruction for the knowledge graph construction tool can be initiated on the terminal device to start the knowledge graph construction tool.

[0067] Correspondingly, the terminal device starts and displays the user the operation interface of the knowledge graph construction tool. Figure 2 As shown. Then, the user can use the above-mentioned knowledge graph construction tool to process the corresponding target source data through the terminal device to construct a target knowledge graph that meets the business needs.

[0068] The operating interface of the knowledge graph construction tool displayed includes at least structures such as a source data import interface, a first processing interface, and a second processing interface.

[0069] In the specific implementation, first, the user can use the source data import interface according to the relevant instructions on the operation interface, select and import a batch of data to be processed as the target source data through various import methods such as local file upload, HDFS file import, database table import, and third-party data access.

[0070] When importing the target source data, users can also customize the import method. For example, they can choose to use the source data import interface to display preview information of the data to be imported during the process of importing the target source data. The preview information can be a data volume parameter (for example, the number of rows, columns, total number, etc.) or a data content parameter (for example, data content keywords, preview of the first few rows of data, data name, etc.). In this way, when using the source data import interface to import the target source data, users can perform the import operation more accurately based on the preview information to avoid import errors.

[0071] Next, the user may use the first processing interface to set definition parameters for data objects and / or definition parameters for data relationships in the target knowledge graph to be constructed.

[0072] Specifically, the user can set the name, attribute, type, etc. of the data object (or entity object) in the first processing interface as the definition parameters of the data object (for example, the customer's name, the customer's account, the customer's company, etc.). The user can also set the name, attribute, type, etc. of the data relationship in the first processing interface as the definition parameters of the data relationship (for example, the ownership relationship between data objects, the transfer relationship between data objects, the debt relationship between data objects, etc.).

[0073] Correspondingly, the terminal device can receive the definition parameters of the data object and / or the definition parameters of the data relationship set by the user through the above-mentioned first processing interface, and generate a definition parameter file about the target knowledge graph based on the above-mentioned definition parameters.

[0074] Through the above method, users can use the first processing interface to generate a definition parameter file of the target knowledge graph based on actual business needs, and flexibly complete the construction of the graph structure of the target knowledge graph.

[0075] Of course, it should be noted that, in the specific implementation, the user may first input specific business requirements in the first processing interface; and then the terminal device may automatically generate a definition parameter file for the target knowledge graph based on the business requirements and the data characteristics of the target source data.

[0076] Then, the user can use the second processing interface to efficiently and easily construct a target knowledge extraction unit that matches the target source data and meets business needs according to the preset construction rules and the data structure type of the target source data. The above-mentioned target knowledge extraction unit is used to process the target source data to extract the entity relationship file used to generate the target knowledge graph. The entity relationship file can specifically include multiple triple data groups extracted based on the target source data. Each triple data group can include at least two data objects and one data relationship, wherein two data objects in the same triple data group can be connected by a data relationship.

[0077] For details, please refer to Figure 2 As shown, the second processing interface may specifically include: a menu bar, a main canvas, a parameter configuration bar and other structures. Specifically, in the above menu bar, a data source operator selection box, a data processing structure selection box, and an identification termination operator selection box may be displayed. In the above parameter configuration bar, a parameter configuration interface for the data source operator, the data processing structure, and the identification termination operator may be specifically provided.

[0078] Furthermore, the above-mentioned data source operator selection box may specifically include a plurality of preset data source operators for user selection, such as the DATAS operator for structured data, the DATAU operator for unstructured data, etc. The above-mentioned data processing structure selection box may specifically include a plurality of preset data processing structures for user selection, such as a plurality of preset data processing operators matching structured data (including: SQL operator, HIVE operator, SPARK operator, etc.), pre-trained preset triple extraction models matching unstructured data or semi-structured data, etc. The above-mentioned identification termination operator selection box may specifically include a plurality of preset identification termination operators for user selection, such as the MDATAS operator, etc.

[0079] Specifically, the user can, based on the preset construction rules, combined with specific business needs and the data structure type of the target source data, first select the matching preset data source operator, preset data processing structure, and preset identification termination operator from the menu bar through the above-mentioned second processing interface; at the same time, the parameter configuration bar can be used to select the selected preset data source operator, preset data processing structure, and preset identification termination operator to obtain the corresponding target source operator, target data processing structure, and target identification termination operator; and then combine the above-mentioned target source operator, target data processing structure, and target identification termination operator through the main canvas to obtain a target knowledge extraction unit that meets the user's personalized business needs and matches the target source data.

[0080] For example, taking the target source data as the transaction data flow record of XX Bank's customers in 2020 as an example, first, based on the preset construction rules, considering that the target source data is structured data imported through the database table, you can choose to use the DATAS operator for structured data and configure the corresponding import parameters to customize the import method of the target source data and obtain the corresponding target source operator. At the same time, you can also choose to use the SQL operator suitable for processing structured data and configure the corresponding processing logic to customize the knowledge extraction method of the target source data to obtain the corresponding target data processing structure. Then, considering the specific business needs, you can choose to use the MDATAS operator and configure the corresponding extraction parameters (for example, the identification information of the data object to be extracted, the identification information of the data relationship to be extracted, etc.) to customize the extracted data objects and data relationships to obtain the corresponding target identification termination operator.

[0081] Next, the user can drag the above-mentioned target source operator, target data processing structure, and target identification termination operator into the main canvas, and arrange them in the order of target source operator, target data processing structure, and target identification termination operator; then use connecting lines to connect the target source operator and the target data processing structure, as well as the target data processing structure and the target identification termination operator, to complete the combination and obtain a target knowledge extraction unit that meets the user's business needs and matches the target source data.

[0082] Through the above method, the user can use the second processing interface to efficiently and conveniently construct a target knowledge extraction unit that meets the requirements based on actual business needs.

[0083] Of course, it should be noted that, in specific implementation, the above process can also be that the terminal device automatically generates the above target knowledge extraction unit based on preset construction rules, according to the data structure type of the target source data and specific business needs.

[0084] After obtaining the target knowledge extraction unit, the user can perform corresponding operations in the knowledge graph construction tool (for example, click the confirmation run icon) to initiate a run instruction. The terminal device can respond to the run instruction initiated by the user, based on the instruction program in the knowledge graph construction tool, call the target instruction extraction unit to process the target source data, so as to efficiently extract the triple data group used to construct the target knowledge graph from the target source data, and then the corresponding entity relationship file.

[0085] Furthermore, the terminal device may generate a target knowledge graph associated with the target data source by performing data mapping based on the instruction program in the knowledge graph construction tool and the entity relationship file and the definition parameter file.

[0086] Through the above method, users only need to perform simple operations to efficiently and accurately build a target knowledge graph that meets diverse business needs.

[0087] After obtaining the target knowledge graph, the terminal device can display the target knowledge graph to the user so that the user can query and use the target knowledge graph.

[0088] Users can also use knowledge graph construction tools to modify, edit, and name the target knowledge graph through terminal devices.

[0089] In addition, the terminal device can also set a corresponding target identifier for the target knowledge graph (for example, the generation number or name of the target knowledge graph can be determined as the target identifier corresponding to the target knowledge graph); and send the target knowledge graph carrying the target identifier to the server. Correspondingly, the server can store the received target knowledge graph carrying the target identifier in the graph database.

[0090] Later, when the user needs to query the target knowledge graph again, the user can generate and send a relevant target query statement to the server through the terminal device, wherein the target query statement carries at least the target identifier.

[0091] Correspondingly, the server receives the target query statement, searches the graph database according to the target identifier carried in the target query statement, and finds the target knowledge graph that the user indicates to query. Then, the server can respond to the target query statement, perform specific query operations on the target knowledge graph, obtain the corresponding query results, and then feed back the query results to the terminal device.

[0092] The terminal device receives and displays the above query results to the user.

[0093] In this way, the terminal device can efficiently and conveniently complete the query of the target knowledge graph and obtain the required query results. Furthermore, the terminal device can perform further data processing based on the query results.

[0094] For example, the terminal device can view the flow diagram of the target customer's transaction data, further analyze whether there is any abnormality in the flow of the target customer's transaction data, and then use this as a basis to determine whether the target customer is at risk of illegal transactions (for example, money laundering risk, gambling risk, etc.). In this way, customers with illegal transaction risks can be identified efficiently and accurately.

[0095] See also Figure 3 As shown, the embodiment of this specification provides a method for constructing a knowledge graph. Among them, when the method is implemented specifically, it can include the following contents.

[0096] S301: Acquire target source data.

[0097] S302: Determine the data structure type of the target source data.

[0098] S303: Constructing a target knowledge extraction unit matching the target source data according to a preset construction rule and the data structure type of the target source data.

[0099] S304: Call the target knowledge extraction unit to process the target source data to obtain an entity relationship file that meets the requirements; wherein the entity relationship file contains multiple triple data groups; the triple data group includes at least two data objects connected by a data relationship.

[0100] S305: Construct a target knowledge graph associated with the target source data based on the entity relationship file.

[0101] Through the above embodiments, user-side operations can be effectively simplified and the difficulty of building a knowledge graph can be reduced, so that users can more efficiently and accurately build a target knowledge graph that meets the user's diverse business needs.

[0102] In some embodiments, the target source data may specifically refer to the source data used to generate the target knowledge graph required by the user. Corresponding to different application scenarios and different business needs, the target source data may specifically be data of different contents. Specifically, for example, in the scenario of customer transaction risk prediction, the target source data may specifically be a flow record of the customer's transaction data (e.g., asset data, financial data, etc.). For another example, in the scenario of verifying the relationship between historical figures, the target source data may specifically be communication records between different historical figures.

[0103] In some embodiments, the target source data may include data of multiple different data structure types. Specifically, the data structure type of the target source data may include at least one of the following: structured data, unstructured data, semi-structured data, etc.

[0104] The structured data mentioned above may specifically refer to data that meets a preset data format (e.g., a key-value pair format, etc.). Usually, for a certain structured data, according to the preset data format corresponding to the data, the specific attributes of the different data contained in the data can be relatively directly determined. For example, for a data that meets the key-value pair format, it is relatively directly possible to determine which data in the data is the key value and which data is the value value.

[0105] The above-mentioned semi-structured data may specifically refer to data that does not meet the preset data format but still meets certain other conventional formats (e.g., table format, etc.). Generally, for a semi-structured data, although it is not possible to directly determine the specific attributes of the different data contained in the data as with structured data; however, in combination with the corresponding conventional format, through certain semantic analysis processing, the specific attributes of the different data contained in the data can also be determined.

[0106] The above unstructured data may specifically refer to data that does not meet the preset data format and does not meet certain conventional formats, such as a text message in a transaction order, etc. Usually, semantic analysis is required for unstructured data to determine the specific attributes of the different data contained in the data.

[0107] Through the above embodiments, the knowledge graph construction method provided in the embodiments of this specification can be extended to process target source data of various different data structure types to meet the diverse business needs of users.

[0108] In some embodiments, when specifically acquiring the target source data, for structured data, the target source data can be acquired by any of the following acquisition methods: acquiring the target source data by uploading a local file; acquiring the target source data by importing a HDFS file; acquiring the target source data by importing a database table, etc.

[0109] For semi-structured data and unstructured data, the target source data can be obtained by any of the following acquisition methods: obtaining the target source data by uploading local files; obtaining the target source data by accessing data provided by a third party; obtaining the target source data by receiving data transmitted by other distributed clusters, etc.

[0110] In some embodiments, in order to obtain the target source data more efficiently, when obtaining the target source data through HDFS file import, the characteristics of HDFS can be used in advance to record the file path of the target source data (corresponding to HDFS metadata) in the database; when the target source data needs to be obtained, the database can be queried to obtain and use the above file path to directly access and obtain the corresponding target source data. This can avoid multiple landings in the data acquisition process and improve the efficiency of obtaining the target source data.

[0111] In some embodiments, during specific implementation, the data structure type of the target source data may be detected and determined based on the data structure characteristics of the target source data.

[0112] In some embodiments, in order to better balance the processing of target source data of multiple different data structure types, the data structure type of the target source data can be determined first; then, based on the data structure type of the target source data, the target source data of different data structure types can be distinguished, and the target source data of different data structure types can be processed in a matching manner to construct a target knowledge graph that meets the user's business needs.

[0113] Specifically, we can first distinguish two major categories: the first category (including structured data) and the second category (including unstructured data and semi-structured data); then, for the above two categories, we construct matching target knowledge extraction units according to preset construction rules; and then we can use the matching target knowledge extraction units to process the target source data to construct the corresponding target knowledge graph.

[0114] In some embodiments, the target knowledge extraction unit matching the target source data is constructed according to the preset construction rules and the data structure type of the target source data. When implemented specifically, the following contents may be included:

[0115] S1: According to a preset construction rule, a target source operator corresponding to the target source data is screened out from a plurality of preset data source operators; wherein the target source operator is used to connect the target source data to a target knowledge extraction unit;

[0116] S2: Determine a matching target data processing structure according to the data structure type of the target source data; wherein the target data processing structure is used to process the target source data to obtain a plurality of triple data groups;

[0117] S3: Determine and configure a target identification termination operator; wherein the target identification termination operator is used to extract a triple data group that meets the requirements from a plurality of triple data groups output by the target data processing structure to obtain a corresponding entity relationship file;

[0118] S4: combining the target source operator, the target data processing structure and the target identification termination operator to obtain a target knowledge extraction unit that matches the target data source.

[0119] Through the above embodiments, based on preset construction rules and in combination with data features such as the data structure type of the target source data, a target knowledge extraction unit that matches the target source data and meets the user's business needs can be accurately established.

[0120] In some embodiments, the above-mentioned determination of a matching target data processing structure based on the data structure type of the target source data may include the following contents during specific implementation: when it is determined that the data structure type of the target source data is structured data, selecting an initial processing operator from a plurality of preset data processing operators; configuring the initial processing operator accordingly to obtain a target processing operator; and determining the target processing operator as a matching target data processing structure.

[0121] Through the above embodiments, based on preset construction rules, a more targeted and more matching target data processing structure can be determined for target source data whose data structure type is structured data.

[0122] In some embodiments, a preset data processing operator that meets the requirements can be screened out from multiple preset data processing operators as an initial processing operator according to the user's business requirements, data characteristics of the target source data, and the user's preferred programming language, etc. Further, the processing logic of the initial processing operator can be configured according to the business requirements, so as to obtain a more matching target data processing structure.

[0123] In some embodiments, the preset data processing operator may specifically include at least one of the following: a SQL operator, a HIVE operator, a SPARK operator, etc.

[0124] Of course, it should be noted that the preset data processing operators listed above are only schematic illustrations. In specific implementation, other types of data processing operators may be introduced as preset data processing operators according to specific circumstances and the programming language used.

[0125] Through the above embodiments, for target source data whose data structure type is structured data, a variety of optional preset data processing operators can be prepared and provided to obtain a target data processing structure that meets the requirements, thereby obtaining a target data processing structure with a relatively higher matching degree and more accurate.

[0126] In some embodiments, the above-mentioned determination of a matching target data processing structure based on the data structure type of the target source data may also include, when it is determined that the data structure type of the target source data is unstructured data or semi-structured data, determining a preset triple extraction model as a matching target data processing structure.

[0127] Through the above embodiments, based on the preset construction rules, a more targeted and matching target data processing structure can be determined for target source data whose data structure type is unstructured data or semi-structured data.

[0128] In some embodiments, the preset triple extraction model may specifically refer to a pre-trained model that can extract corresponding triple data groups from text data based on semantic recognition. The triple data group may specifically include two data objects connected by a data relationship.

[0129] In some embodiments, in the scenario of predicting transaction risks of customers, the above data objects may specifically be the name of the customer, the customer's account, or the customer's holding company, etc. The above data relationships may specifically be the transfer relationship between data objects, the interest ownership relationship between data objects, the debt relationship between data objects, etc. Of course, the above-listed data objects and data relationships are only a schematic illustration. According to specific application scenarios and business requirements, the above data objects and data relationships may also be data of other contents.

[0130] In some embodiments, before specific implementation, a preset triple extraction model can be trained in the following manner: obtain sample text data; mark two data objects that have a data relationship in the sample text data to obtain the marked sample text data; use the marked sample text data to perform model training to obtain the preset triple extraction model.

[0131] In some embodiments, after determining the data structure type of the target source data, the method may also include the following contents when implemented: screening out a recommended knowledge extraction unit from a plurality of preset knowledge extraction units based on the data structure type of the target source data; displaying the recommended knowledge extraction unit to the user; and determining the recommended knowledge extraction unit selected by the user as the target knowledge extraction unit.

[0132] Through the above embodiments, before specific implementation, multiple preset knowledge extraction units can be pre-configured according to historical processing records for a variety of relatively common target source data and relatively common business needs; during specific implementation, the preset knowledge extraction units that match the data structure type of the target source data can be screened out from multiple preset knowledge extraction units as recommended knowledge extraction units for user selection according to the data structure type of the target source data; accordingly, the user only needs to select the recommended knowledge extraction unit that meets his or her business needs from multiple existing recommended knowledge extraction units as the target knowledge extraction unit according to specific business needs. In this way, the target knowledge extraction unit that meets the user's business needs can be obtained more efficiently and conveniently.

[0133] In some embodiments, after the target knowledge extraction unit is constructed, the target knowledge extraction unit can be called to process the target source data to extract triple data groups from the target source data; then, triple data groups associated with the user's business needs are screened out from the triple data groups to construct an entity relationship file that meets the requirements. Specifically, the entity relationship file can include multiple triple data groups associated with the user's business needs.

[0134] In some embodiments, according to the entity relationship file, the above-mentioned construction of the target knowledge graph associated with the target source data may include the following contents during specific implementation:

[0135] S1: Obtain a definition parameter file about a target knowledge graph; wherein the definition parameter file includes: definition parameters of data objects and / or definition parameters of data relationships;

[0136] S2: Generate a target knowledge graph associated with the target data source by performing data mapping according to the entity relationship file and the definition parameter file.

[0137] Through the above-mentioned embodiments, entity relationship files and definition parameter files can be used to efficiently and accurately construct the target knowledge graph required by the user through data mapping.

[0138] In some embodiments, the target knowledge graph may be a graph including multiple nodes and connecting edges, wherein each node corresponds to a data object, each connecting edge corresponds to at least one data relationship, and the nodes are connected via connecting edges.

[0139] In some embodiments, when performing data mapping according to the entity relationship file and the definition parameter file, the data object in the definition parameter file can be mapped into a node according to the entity relationship file, and the data relationship can be mapped into a connecting edge and converted into corresponding graph data, so that the target knowledge graph can be constructed.

[0140] In some embodiments, when constructing the target knowledge graph, the attribute information of the nodes and / or the attribute information of the connecting edges can also be determined based on the entity relationship file and the definition parameter file; and the attribute information of the corresponding nodes and / or connecting edges in the knowledge graph can be annotated, so as to obtain a target knowledge graph with relatively richer data content and relatively better effects.

[0141] In some embodiments, the definition parameter file may further specifically include: index definition parameters; accordingly, in the process of constructing a target knowledge graph associated with the target source data according to the entity relationship file, the method, when implemented specifically, may further include: constructing a target query index for the target knowledge graph according to the index definition parameters, using the definition parameters of the data object and / or the definition parameters of the data relationship.

[0142] Through the above embodiment, while constructing the target knowledge graph, it is also possible to construct a target query index for the target knowledge graph based on the index definition parameters in the definition parameter file, so that the target query index can be used later to more efficiently use and query the target knowledge graph.

[0143] In some embodiments, during specific implementation, the constructed target knowledge graph can be stored in a graph database together with the corresponding target query index for subsequent use.

[0144] In some embodiments, considering that when batch-building multiple target knowledge graphs with large target source data volumes, a large amount of data processing resources are often consumed, which can easily form a large processing load on the system (or server, terminal device, etc.), affecting the stability of the overall operation of the system. Therefore, when batch-building multiple target knowledge graphs, the system can also be set to first estimate whether the amount of data processing required to build each target knowledge graph is greater than a preset threshold processing amount. Among them, the above-mentioned preset threshold processing amount can be specifically determined based on the overall processing performance of the system.

[0145] When it is determined that the amount of data processing required to construct the target knowledge graph is less than the preset threshold processing amount, the system can load the data normally, process and construct the corresponding target knowledge graph.

[0146] When it is determined that the amount of data processing required to build the target knowledge graph is greater than or equal to the preset threshold processing, the system can suspend loading data and the construction of the target knowledge graph, and prompt the user who initiated the target knowledge graph that the construction of the target knowledge graph needs to be approved first, and can only be executed normally if the approval is passed. In addition, the system can also monitor the load status of the system in real time. When it is determined that the load status of the system allows the construction and processing of the target knowledge graph, it will resume loading data and perform the corresponding construction of the target knowledge graph. This can protect the overall stable and reliable operation of the system.

[0147] In some embodiments, after constructing a target knowledge graph associated with the target source data according to the entity relationship file, the method may also include the following contents when implemented: receiving a target query statement; wherein the target query statement carries at least a target identifier of the target knowledge graph; searching a graph database according to the target identifier to determine the target knowledge graph; in response to the target query statement, performing a query operation on the target knowledge graph to obtain a corresponding query result; and feeding back the query result.

[0148] Through the above embodiments, it is possible to respond to the target query statement initiated by the user, efficiently find the corresponding target knowledge graph from the graph database for query operation, and promptly feedback the relevant query results to the user.

[0149] In some embodiments, after determining the target knowledge graph, it is also possible to detect whether the target query index of the target knowledge graph is stored in the graph database; when the target query index is detected, the target query statement can be responded to, and combined with the target query index, query operations can be performed on the target knowledge graph more efficiently and accurately, thereby further improving query efficiency and improving the user's query experience.

[0150] In some embodiments, the target source data includes a flow record of a customer's transaction data; correspondingly, the query result includes a flow diagram of the target customer's transaction data.

[0151] Specifically, for example, in a scenario of predicting a customer's transaction risk, the transaction data may be capital data, and the query result may be a flow chart of the target customer's capital data. In specific implementation, the flow chart of the target customer's capital data may be used to analyze whether there is an abnormality in the flow of the target customer's capital data, and further determine whether the target customer has corresponding transaction risks (e.g., money laundering risk, gambling risk, fraud risk, etc.).

[0152] Through the above embodiments, the knowledge graph construction method provided in the embodiments of this specification can be better applied to the prediction scenario of customer transaction risks, so as to use the constructed target knowledge graph to accurately and efficiently predict whether the target customer has corresponding transaction risks.

[0153] As can be seen from the above, the knowledge graph construction method provided in the embodiment of this specification can first determine the data structure type of the target source data to be processed based on the knowledge graph construction method; then, according to the preset construction rules and the data structure type of the target source data, construct a target knowledge extraction unit that matches the target source data; further, the above target knowledge extraction unit can be called to specifically process the target source data to obtain an entity relationship file that meets the requirements and contains multiple triple data groups; then, according to the above entity relationship file, construct a target knowledge graph associated with the target source data. This can effectively simplify user-side operations and reduce the difficulty of building a knowledge graph, so that users can efficiently and accurately build a knowledge graph that meets diverse business needs and has better effects.

[0154] See also Figure 2 As shown, the embodiment of this specification also provides a knowledge graph construction tool. The knowledge graph construction tool may at least include: a source data import interface, a first processing interface, and a second processing interface; wherein,

[0155] The source data import interface can be specifically used to support users to import target source data;

[0156] The first processing interface may be specifically used to support a user in setting definition parameters of data objects and / or definition parameters of data relationships in a target knowledge graph, so as to generate a definition parameter file for the target knowledge graph;

[0157] The second processing interface can be specifically used to support the user to determine and combine matching target source operators, target data processing structures, and identification termination operators according to preset construction rules to obtain a target knowledge extraction unit that matches the target source data;

[0158] The knowledge graph construction tool is used to call the target knowledge extraction unit to process the target source data to obtain the corresponding entity relationship file; and based on the entity relationship file and the definition parameter file, generate a target knowledge graph associated with the target data source by performing data mapping.

[0159] Through the above embodiments, users can use the above knowledge graph construction tools to efficiently and conveniently achieve one-stop construction of target knowledge graphs that meet diverse business needs, thereby effectively simplifying user-side operations, reducing the difficulty of knowledge graph construction, and improving user experience.

[0160] An embodiment of the present specification also provides a server, including a processor and a memory for storing processor executable instructions, wherein the processor can perform the following steps according to the instructions when implemented: obtaining target source data; determining the data structure type of the target source data; constructing a target knowledge extraction unit matching the target source data according to preset construction rules and the data structure type of the target source data; calling the target knowledge extraction unit to process the target source data to obtain an entity relationship file that meets the requirements; wherein the entity relationship file contains multiple triple data groups; the triple data group includes at least two data objects connected by a data relationship; and constructing a target knowledge graph associated with the target source data according to the entity relationship file.

[0161] In order to complete the above instructions more accurately, refer to Figure 4 As shown, the embodiment of this specification also provides another specific server, wherein the server includes a network communication port 401, a processor 402 and a memory 403, and the above structures are connected through internal cables so that each structure can perform specific data interaction.

[0162] The network communication port 401 can be used to obtain target source data.

[0163] The processor 402 can be specifically used to determine the data structure type of the target source data; construct a target knowledge extraction unit matching the target source data according to a preset construction rule and the data structure type of the target source data; call the target knowledge extraction unit to process the target source data to obtain an entity relationship file that meets the requirements; wherein the entity relationship file contains multiple triple data groups; the triple data group includes at least two data objects connected by a data relationship; and construct a target knowledge graph associated with the target source data according to the entity relationship file.

[0164] The memory 403 may be specifically used to store corresponding instruction programs.

[0165] In this embodiment, the network communication port 401 can be a virtual port that is bound to different communication protocols so that different data can be sent or received. For example, the network communication port can be a port responsible for web data communication, a port responsible for FTP data communication, or a port responsible for email data communication. In addition, the network communication port can also be a physical communication interface or communication chip. For example, it can be a wireless mobile network communication chip, such as GSM, CDMA, etc.; it can also be a Wifi chip; it can also be a Bluetooth chip.

[0166] In this embodiment, the processor 402 may be implemented in any appropriate manner. For example, the processor may take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (such as software or firmware) executable by the (micro)processor, a logic gate, a switch, an application specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller, etc. This specification does not limit this.

[0167] In this embodiment, the memory 403 may include multiple levels. In a digital system, anything that can store binary data can be a memory; in an integrated circuit, a circuit with a storage function that has no physical form is also called a memory, such as RAM, FIFO, etc.; in a system, a storage device with a physical form is also called a memory, such as a memory stick, TF card, etc.

[0168] The embodiments of this specification also provide a computer storage medium based on the above-mentioned knowledge graph construction method, wherein the computer storage medium stores computer program instructions, which, when executed, implement: obtaining target source data; determining the data structure type of the target source data; constructing a target knowledge extraction unit matching the target source data according to preset construction rules and the data structure type of the target source data; calling the target knowledge extraction unit to process the target source data to obtain an entity relationship file that meets the requirements; wherein the entity relationship file contains multiple triple data groups; the triple data groups include at least two data objects connected by a data relationship; and constructing a target knowledge graph associated with the target source data according to the entity relationship file.

[0169] In this embodiment, the storage medium includes, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a cache, a hard disk (HDD), or a memory card. The memory may be used to store computer program instructions. The network communication unit may be an interface for network connection communication set in accordance with the standard specified by the communication protocol.

[0170] In this embodiment, the functions and effects specifically implemented by the program instructions stored in the computer storage medium can be explained in comparison with other implementations and will not be described in detail here.

[0171] See also Figure 5As shown, at the software level, the embodiment of this specification also provides a knowledge graph construction device, which may specifically include the following structural modules:

[0172] The acquisition module 501 may be specifically used to acquire target source data;

[0173] A determination module 502 may be specifically used to determine a data structure type of target source data;

[0174] The first construction module 503 may be specifically used to construct a target knowledge extraction unit matching the target source data according to a preset construction rule and a data structure type of the target source data;

[0175] The calling module 504 may be specifically used to call the target knowledge extraction unit to process the target source data to obtain an entity relationship file that meets the requirements; wherein the entity relationship file includes a plurality of triple data groups; the triple data group includes at least two data objects connected by a data relationship;

[0176] The second construction module 505 can be specifically used to construct a target knowledge graph associated with the target source data based on the entity relationship file.

[0177] It should be noted that the units, devices or modules described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. For the convenience of description, the above devices are described separately by functions divided into various modules. Of course, when implementing this specification, the functions of each module can be implemented in the same or more software and / or hardware, or the modules that implement the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0178] It can be seen from the above that the knowledge graph construction device provided in the embodiments of this specification can effectively simplify user-side operations and reduce the difficulty of constructing the knowledge graph, so that users can efficiently and accurately construct a knowledge graph that meets diverse business needs and has better effects.

[0179] In a specific scenario example, the knowledge graph construction method provided in the embodiment of this specification can be applied to import data to construct a corresponding knowledge graph. Figure 6 As shown, the following steps are included.

[0180] Step 1: Import various types of data (eg, target source data of multiple different data structure types) into the system.

[0181] Step 2: Build a graph structure according to scenario requirements (for example, generate a corresponding definition parameter file).

[0182] In this scenario example, the construction of the graph structure may include: defining the entity information contained in the graph (for example, definition parameters of data objects), relationship information (for example, definition parameters of data relationships), attribute information of entity relationships, and index information. Specifically, for example, the entity VERTEX may be defined including but not limited to the type label of VERTEX and various attributes of VERTEX; the relationship EDGE may be defined including but not limited to the type label of EDGE, the starting point type of EDGE, and various attributes of EDGE. For scenarios with specific query requirements, the efficiency of subsequent queries can be improved by constructing an index (for example, a target query index), which can be specifically constructed on entities, relationships, or attributes, and a combination of the three. For example, a point index can be constructed on VERTEX to improve the query efficiency of scenarios with query requirements starting with a point.

[0183] Step 3: Structural processing of each type of data (to obtain the corresponding entity relationship file).

[0184] In this scenario example, for structured data, first select the structured data DATAS operator (for example, the target source operator) to be processed in the main sidebar (in the second processing interface) and drag it into the main canvas. Secondly, select the data processing operator in the main sidebar, such as the SQL operator, fill in the specific processing logic of the SQL operator in the configuration sidebar (to obtain the target data processing structure), and click Run Operator. After the operation is successful, select the identification termination operator, fill in the termination operator name MDATAS (to obtain the target identification termination operator), and click Run Operator until the operation is successful.

[0185] For unstructured data or semi-structured data, similarly, first drag in the unstructured data DATAU, then select the trained model operator (for example, the preset triple extraction model), click Run Prediction, and after the operator runs successfully, select the marker to terminate the operator MDATAU until it runs successfully.

[0186] Step 4: Perform knowledge mapping between ontology and data.

[0187] In this scenario example, first, you can select the data source to be used as a candidate data source in the full data, and then select the ontology model to be mapped. Then, select an entity VERTEX in the ontology model, and select the data source MDATAS corresponding to VERTEX. Finally, for each attribute of VERTEX, select a field in MDATAS to map it one by one. And so on, map all entity relationships with entity relationship data source files.

[0188] Step 5: Import the graph structure and data into the database (generate the corresponding knowledge graph and store it in the graph database).

[0189] In this scenario embodiment, when generating and storing the knowledge graph, some configuration information of the graph can also be filled in, including but not limited to the graph name, etc. Then click Graph Import to import the ontology model and data into the knowledge graph in batches.

[0190] Step 6: Visualize the graph through the visualization module.

[0191] In this scenario example, when the user needs to query graph data, he can fill in and send the corresponding query statement to query the graph data and visualize the query results.

[0192] Through the above scenario examples, it is verified that the knowledge graph construction method provided in the embodiments of this specification is a one-stop, relatively simple and efficient method, which has the following advantages: First, it provides a simple and efficient graph construction and visualization display platform for business personnel without technical background who use knowledge graph technology for analysis and exploration, and proposes usability optimization for each link of the construction, which solves the disadvantage that the graph has a large number of available scenarios but a high technical threshold; second, it comprehensively considers and summarizes several indispensable aspects in the graph construction process to form a simple, easy-to-use and low-threshold one-stop graph construction system. The graph construction processes refer to ETL for module splitting. Compared with the existing processes in the industry, the various processes of graph construction are displayed more clearly, which is in line with the user's thinking and cognitive logic, and provides new ideas for the creation of graph construction tools; third, for the knowledge extraction module, a DAG-based data processing process is proposed, which is more powerful and easy to use than the common rule model.

[0193] Although the present specification provides method operation steps as described in the embodiments or flow charts, more or less operation steps may be included based on conventional or non-creative means. The order of steps listed in the embodiments is only one way of executing the order of many steps, and does not represent a unique execution order. When the device or client product in practice is executed, it can be executed in sequence or in parallel according to the method shown in the embodiments or the drawings (for example, a parallel processor or a multi-threaded processing environment, or even a distributed data processing environment). The term "include", "include" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, product or device including a series of elements includes not only those elements, but also includes other elements that are not explicitly listed, or also includes elements inherent to such a process, method, product or device. In the absence of more restrictions, it is not excluded that there are other identical or equivalent elements in the process, method, product or device including the elements. The first, second, etc. words are used to represent the name, and do not represent any particular order.

[0194] Those skilled in the art also know that, in addition to implementing the controller in a purely computer-readable program code, the controller can be made to implement the same function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, such a controller can be considered as a hardware component, and the devices for implementing various functions included therein can also be considered as structures within the hardware component. Or even, the devices for implementing various functions can be considered as both software modules for implementing the method and structures within the hardware component.

[0195] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, classes, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media including storage devices.

[0196] Through the description of the above embodiments, it can be known that those skilled in the art can clearly understand that the present specification can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present specification can essentially be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a mobile terminal, a server, or a network device, etc.) to execute the methods described in each embodiment of the present specification or some parts of the embodiments.

[0197] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. This specification can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld devices or portable devices, tablet devices, multi-processor systems, microprocessor-based systems, set-top boxes, programmable electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc.

[0198] Although the present specification is described through embodiments, those skilled in the art will appreciate that there are many modifications and changes to the present specification without departing from the spirit of the present specification, and it is intended that the appended claims include these modifications and changes without departing from the spirit of the present specification.

Claims

1. A method for constructing a knowledge graph, characterized in that: include: Get target source data; Determine the data structure type of the target source data; According to the preset construction rules and the data structure type of the target source data, a target knowledge extraction unit matching the target source data is constructed; including: according to the preset construction rules, a target source operator corresponding to the target source data is screened out from a plurality of preset data source operators; wherein the target source operator is used to connect the target source data to the target knowledge extraction unit; according to the data structure type of the target source data, a matching target data processing structure is determined; wherein the target data processing structure is used to process the target source data to obtain a plurality of triple data groups; a target identification termination operator is determined and configured; wherein the target identification termination operator is used to extract a triple data group that meets the requirements from a plurality of triple data groups output by the target data processing structure to obtain a corresponding entity relationship file; the target source operator, the target data processing structure and the target identification termination operator are combined to obtain a target knowledge extraction unit matching the target source data; Calling the target knowledge extraction unit to process the target source data to obtain an entity relationship file that meets the requirements; wherein the entity relationship file includes a plurality of triple data groups; the triple data group includes at least two data objects connected by a data relationship; Constructing a target knowledge graph associated with the target source data according to the entity relationship file; The method further includes: when it is determined that the data structure type of the target source data is structured data, based on the business requirements of the user, the data characteristics of the target source data, and the programming language preferred by the user, selecting a preset data processing operator that meets the requirements from multiple preset data processing operators as an initial processing operator; and configuring the processing logic of the initial processing operator according to the business requirements to obtain a matching target data processing structure; And, when it is determined that the data structure type of the target source data is unstructured data or semi-structured data, the preset triple extraction model is determined as the matching target data processing structure; wherein, the preset triple extraction model is a pre-trained model that can extract corresponding triple data groups from text data based on semantic recognition.

2. The method according to claim 1, characterized in that The data structure type of the target source data includes at least one of the following: structured data, unstructured data, and semi-structured data.

3. The method according to claim 1, characterized in that The preset data processing operator includes at least one of the following: SQL operator, HIVE operator, SPARK operator.

4. The method according to claim 1, characterized in that: After determining the data structure type of the target source data, the method further includes: According to the data structure type of the target source data, a recommended knowledge extraction unit is selected from a plurality of preset knowledge extraction units; Displaying the recommended knowledge extraction unit to a user; The recommended knowledge extraction unit selected by the user is determined as the target knowledge extraction unit.

5. The method according to claim 1, characterized in that According to the entity relationship file, a target knowledge graph associated with the target source data is constructed, including: Obtain a definition parameter file about the target knowledge graph; wherein the definition parameter file includes: definition parameters of data objects and / or definition parameters of data relationships; According to the entity relationship file and the definition parameter file, a target knowledge graph associated with the target source data is generated by performing data mapping.

6. The method according to claim 5, characterized in that The definition parameter file also includes index definition parameters; Accordingly, in the process of constructing a target knowledge graph associated with the target source data according to the entity relationship file, the method further includes: According to the index definition parameters, a target query index for the target knowledge graph is constructed using definition parameters of the data object and / or definition parameters of the data relationship.

7. The method according to claim 1, characterized in that After constructing a target knowledge graph associated with the target source data according to the entity relationship file, the method further includes: Receive a target query statement; wherein the target query statement carries at least a target identifier of a target knowledge graph; According to the target identifier, searching the graph database to determine the target knowledge graph; In response to the target query statement, a query operation is performed on the target knowledge graph to obtain a corresponding query result; Feedback the query result.

8. The method according to claim 7, characterized in that The target source data includes a flow record of the customer's transaction data; accordingly, the query result includes a flow diagram of the target customer's transaction data.

9. A knowledge graph construction tool, characterized in that: At least: Source data import interface, first processing interface, second processing interface; among them, The source data import interface is used to support users to import target source data; The first processing interface is used to support the user to set definition parameters of data objects and / or definition parameters of data relationships in the target knowledge graph to generate a definition parameter file for the target knowledge graph; The second processing interface is used to support the user to determine and combine matching target source operators, target data processing structures, and identification termination operators according to preset construction rules to obtain a target knowledge extraction unit that matches the target source data; including: according to the preset construction rules, screening out the target source operator corresponding to the target source data from multiple preset data source operators; wherein the target source operator is used to connect the target source data to the target knowledge extraction unit; according to the data structure type of the target source data, determining a matching target data processing structure; wherein the target data processing structure is used to process the target source data to obtain multiple triple data groups; determining and configuring a target identification termination operator; wherein the target identification termination operator is used to extract the triple data groups that meet the requirements from the multiple triple data groups output by the target data processing structure to obtain the corresponding entity relationship file; combining the target A source operator, a target data processing structure and a target identification termination operator are used to obtain a target knowledge extraction unit that matches the target source data; wherein, when it is determined that the data structure type of the target source data is structured data, according to the user's business needs, the data characteristics of the target source data, and the user's preferred programming language, a preset data processing operator that meets the requirements is screened out from multiple preset data processing operators as an initial processing operator; according to business needs, the processing logic of the initial processing operator is configured to obtain a matching target data processing structure; and, when it is determined that the data structure type of the target source data is unstructured data or semi-structured data, a preset triple extraction model is determined as a matching target data processing structure; wherein the preset triple extraction model is a pre-trained model that can extract corresponding triple data groups from text data based on semantic recognition; The knowledge graph construction tool is also used to call the target knowledge extraction unit to process the target source data to obtain the corresponding entity relationship file; and based on the entity relationship file and the definition parameter file, generate a target knowledge graph associated with the target source data by performing data mapping.

10. A knowledge graph construction device, characterized in that: include: An acquisition module is used to acquire target source data; A determination module, used to determine the data structure type of the target source data; A first construction module is used to construct a target knowledge extraction unit matching the target source data according to a preset construction rule and a data structure type of the target source data; A calling module, used for calling the target knowledge extraction unit to process the target source data to obtain an entity relationship file that meets the requirements; wherein the entity relationship file contains a plurality of triple data groups; the triple data group includes at least two data objects connected by a data relationship; A second construction module is used to construct a target knowledge graph associated with the target source data according to the entity relationship file; The first construction module is specifically used to filter out a target source operator corresponding to the target source data from multiple preset data source operators according to a preset construction rule; wherein the target source operator is used to connect the target source data to the target knowledge extraction unit; according to the data structure type of the target source data, determine a matching target data processing structure; wherein the target data processing structure is used to process the target source data to obtain multiple triple data groups; determine and configure a target identification termination operator; wherein the target identification termination operator is used to extract a triple data group that meets the requirements from multiple triple data groups output by the target data processing structure to obtain a corresponding entity relationship file; combine the target source operator, the target data processing structure and the target identification termination operator to obtain a target knowledge extraction unit that matches the target source data; Wherein, when it is determined that the data structure type of the target source data is structured data, a preset data processing operator that meets the requirements is selected from multiple preset data processing operators according to the user's business requirements, the data characteristics of the target source data, and the user's preferred programming language as an initial processing operator; according to the business requirements, the processing logic of the initial processing operator is configured to obtain a matching target data processing structure; And, when it is determined that the data structure type of the target source data is unstructured data or semi-structured data, the preset triple extraction model is determined as the matching target data processing structure; wherein, the preset triple extraction model is a pre-trained model that can extract corresponding triple data groups from text data based on semantic recognition.

11. A server, characterized in that: The method comprises a processor and a memory for storing processor-executable instructions, wherein the processor implements the steps of the method according to any one of claims 1 to 8 when executing the instructions.

12. A computer storage medium, characterized in that: Computer instructions are stored thereon, and when the instructions are executed, the steps of the method according to any one of claims 1 to 8 are implemented.

Images