WO2021190091A1

WO2021190091A1 - Knowledge map construction method and device based on knowledge node belonging degree

Info

Publication number: WO2021190091A1
Application number: PCT/CN2021/071056
Authority: WO
Inventors: 董润华; 徐国强
Original assignee: 深圳壹账通智能科技有限公司
Priority date: 2020-03-26
Filing date: 2021-01-11
Publication date: 2021-09-30
Also published as: CN111444352A

Abstract

A knowledge map construction method and device based on a knowledge node belonging degree. The method comprises: obtaining a target knowledge node and a slave knowledge node of the target knowledge node (S110), wherein the target knowledge node and the slave knowledge node are knowledge nodes in a preset field; calculating a belonging degree from the slave knowledge node to the target knowledge node according to a path length from the target knowledge node to the slave knowledge node (S120), wherein the path length is determined by the number of intermediate knowledge nodes connected from the target knowledge node to the slave knowledge node; when the belonging degree from the slave knowledge node to the target knowledge node is lower than a threshold, deleting the slave knowledge node, and obtaining the remaining knowledge nodes in the preset field (S130); and constructing a knowledge map according to the remaining knowledge nodes in the preset field (S140). The use of the method can improve the construction efficiency of the knowledge map.

Description

Method and device for constructing knowledge graph based on degree of belonging of knowledge node

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on March 26, 2020, the application number is 202010224365.8, and the invention title is "Knowledge Graph Construction Method and Apparatus Based on the Degree of Knowledge Node Ownership", the entire content of which is incorporated by reference Incorporated in this application.

Technical field

This application relates to the field of computer technology, and in particular to a method and device for constructing a knowledge graph based on the degree of belonging of a knowledge node.

Background technique

Knowledge graph is a large-scale semantic network, which uses entities or concepts as knowledge nodes and is connected by semantic relations. By exploring the associations between entities and integrating semi-structured and unstructured data, knowledge graphs can help machines understand data, explain phenomena, and reason about knowledge, thereby discovering deep relationships and realizing intelligent search and intelligent interaction. At present, it is widely used in banking, insurance, securities, courts, logistics, finance and taxation and other fields. The knowledge graph is composed of multiple knowledge nodes, and multiple knowledge nodes are related to each other, and their interrelationships are represented by the lines in the knowledge graph. The inventor realizes that at present, to construct a knowledge graph of a vertical field, a large amount of manual monitoring is required in the knowledge node acquisition stage to ensure that the acquired knowledge node conforms to the current field. For example, based on the seed map, all the knowledge nodes in the vertical domain are in the seed map, and then according to the knowledge node in the vertical domain, go to the encyclopedia page to search for the next knowledge node associated with it; in the search process, according to each encyclopedia page page (target The tag of the knowledge node) jumps to another encyclopedia webpage (the next knowledge node), and each time the number of jumps increases by 1, it may pull out the content of multiple (for example, 1000) encyclopedia webpages, and whether these contents belong to The seed map needs to be manually deleted, selected and identified.

However, the inventor found that the existing knowledge graphs of vertical fields have knowledge nodes that are not related to the field, which makes it easy to derive wrong conclusions when applying the knowledge graphs.

technical problem

Based on this, it is necessary to address the above technical problems and provide a method, device, computer equipment and storage medium for constructing a knowledge graph based on the degree of belonging of a knowledge node that can improve the efficiency of constructing a knowledge graph.

Technical solutions

A method for constructing a knowledge graph based on the degree of belonging of a knowledge node, the method comprising: obtaining a target knowledge node and a subordinate knowledge node of the target knowledge node; the target knowledge node and the subordinate knowledge node are in a preset domain Knowledge node; according to the path length from the target knowledge node to the subordinate knowledge node, calculate the degree of belonging of the subordinate knowledge node to the target knowledge node; wherein the path length is from the target knowledge node to the The number of intermediate knowledge nodes connected in the subordinate knowledge node is determined; when the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than the threshold, the subordinate knowledge node is deleted, and the information in the preset domain Remaining knowledge nodes; construct a knowledge graph according to the remaining knowledge nodes in the preset domain.

A knowledge graph construction device based on the degree of belonging of a knowledge node, the device comprising: a knowledge node selection module for obtaining a target knowledge node and a subordinate knowledge node of the target knowledge node; the target knowledge node and the subordinate knowledge A node is a knowledge node in a preset field; a degree calculation module for calculating the degree of belonging from the subordinate knowledge node to the target knowledge node according to the path length from the target knowledge node to the subordinate knowledge node; wherein , The path length is determined by the number of intermediate knowledge nodes connected from the target knowledge node to the subordinate knowledge node; the node deletion module is used to determine the degree of belonging of the subordinate knowledge node to the target knowledge node When the value is lower than the threshold, delete the subordinate knowledge node and obtain the remaining knowledge nodes in the preset domain; the knowledge graph construction module is used to construct a knowledge graph based on the remaining knowledge nodes in the preset domain.

A computer device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the method for constructing a knowledge graph based on the degree of belonging of a knowledge node is implemented, and the knowledge based on the degree of belonging of a knowledge node The graph construction method includes the following steps: obtaining a target knowledge node and subordinate knowledge nodes of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain; The path length of the subordinate knowledge node is calculated, and the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated; wherein, the path length is from the target knowledge node to the intermediate knowledge node connected to the subordinate knowledge node The number is determined; when the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than a threshold, delete the subordinate knowledge node, and obtain the remaining knowledge nodes in the preset domain; according to the preset domain The remaining knowledge nodes of the building a knowledge graph.

A computer-readable storage medium having a computer program stored thereon, which when executed by a processor realizes the above-mentioned method for constructing a knowledge graph based on the degree of belonging of a knowledge node, the method for constructing a knowledge graph based on the degree of belonging of a knowledge node includes the following Step: Obtain a target knowledge node and subordinate knowledge nodes of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain; according to the difference between the target knowledge node and the subordinate knowledge node Path length, calculating the degree of belonging of the subordinate knowledge node to the target knowledge node; wherein the path length is determined by the number of intermediate knowledge nodes connected from the target knowledge node to the subordinate knowledge node; When the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than a threshold, delete the subordinate knowledge node, and obtain the remaining knowledge nodes in the preset domain; construct based on the remaining knowledge nodes in the preset domain Knowledge graph.

Beneficial effect

This application calculates the degree of belonging of the subordinate knowledge node to the target knowledge node to eliminate the subordinate knowledge nodes with low relevance, ensuring the accuracy of the constructed knowledge graph, and calculates the degree of belonging to the knowledge entry. Streamlined and improved the efficiency of knowledge graph construction.

Description of the drawings

FIG. 1 is a schematic flowchart of a method for constructing a knowledge graph based on the degree of belonging of a knowledge node in an embodiment.

Fig. 2 is a schematic flowchart of the steps of obtaining a target knowledge node and a subordinate knowledge node of the target knowledge node in an embodiment.

Fig. 3 is a schematic structural diagram of the connection relationship between knowledge nodes and knowledge nodes in a specific embodiment.

Figure 4 is a schematic diagram of the entity and entity attributes of a knowledge entry in a specific embodiment.

Fig. 5 is a structural block diagram of an apparatus for constructing a knowledge graph based on the degree of belonging of a knowledge node in an embodiment.

Fig. 6 is an internal structure diagram of a computer device in an embodiment.

Embodiments of the present invention

In order to make the purpose, technical solutions, and advantages of this application clearer and clearer, the following further describes the application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

The technical solution of this application can be applied to the fields of artificial intelligence, smart city, digital medical care, blockchain and/or big data technology. Optionally, the data involved in this application, such as knowledge node information and/or knowledge graph, can be stored in a database, or can be stored in a blockchain, which is not limited in this application.

In one embodiment, as shown in FIG. 1, a method for constructing a knowledge graph based on the degree of belonging of a knowledge node is provided. In this embodiment, the method is applied to a server for illustration. It is understandable that the method can also be applied. For the terminal, it can also be applied to a system including a terminal and a server, and is realized through the interaction between the terminal and the server. In this embodiment, the method includes the following steps.

Step S110: Obtain a target knowledge node and a subordinate knowledge node of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain.

Among them, the knowledge node is each node that constitutes the knowledge graph, and each knowledge node is connected to each other in the knowledge graph according to the subordination relationship. The knowledge node can be understood as an entity concept, as shown in Figure 3, T1, T2, C1, C2, C3, C4, and C5 are knowledge nodes. The arrows indicate the connection relationship between the knowledge nodes. The current knowledge node points to its subordinate knowledge nodes through the arrow. Knowledge nodes have corresponding knowledge items, and each knowledge item includes the value of the entity and the value of the entity attribute. For example, "currency" in Figure 4 is an entity, and the value of the entity of "currency" and the value of the entity attribute pass through the solid line. The table in the box area 401 is displayed. The elliptical area 402 includes "Chinese name", "foreign name", "function", "essence"... "Affiliation" is the value of the entity, and the dotted area 403 includes "currency"..." "Attribution" is the value of the entity attribute.

Step S120: Calculate the degree of belonging of the subordinate knowledge node to the target knowledge node according to the path length from the target knowledge node to the subordinate knowledge node; wherein, the path length is from the target knowledge node to the The number of intermediate knowledge nodes connected in the subordinate knowledge nodes is determined.

Wherein, the subordinate knowledge node is the next level knowledge node, the next level knowledge node... or the next N level knowledge node of the target knowledge node. The degree of belonging of the subordinate knowledge node to the target knowledge node is determined by the length of the path connecting the subordinate knowledge node to the target knowledge node; of course, the degree of belonging of the subordinate knowledge node to the target knowledge node is determined by the number of paths that can be connected with the subordinate knowledge to the target knowledge node . That is, the degree of belonging is used to characterize the degree of association between the subordinate knowledge node and the target knowledge node. The greater the number of intermediate knowledge nodes connected from the target knowledge node to the subordinate knowledge node, the smaller the degree of belonging. The more connection paths from the target knowledge node to the subordinate knowledge node, the greater the degree of belonging.

For example, the target knowledge node is A, the subordinate knowledge node is D1, and the path from A to D1 is AB-D1, that is, there is a node B between A and D1; another subordinate knowledge node of the target knowledge node A is D2, from A to D1 The path of D2 is ABC-D2, that is, there are two nodes B and C between A and D2; then the degree of belonging of the subordinate knowledge node D1 to the target knowledge node A is greater than the degree of belonging of the subordinate knowledge node D2 to the target knowledge node A. For example, there is one path from target knowledge node A to subordinate knowledge node D1, and there are two paths from target knowledge node A to subordinate knowledge node D2. Then the degree of belonging of subordinate knowledge node D1 to target knowledge node A is less than that of subordinate knowledge node D2 to target knowledge The degree of belonging of node A.

Step S130, when the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than the threshold, delete the subordinate knowledge node, and obtain the remaining knowledge nodes in the preset domain.

The threshold value needs to be set. The higher the threshold value is, the stricter the condition for filtering subordinate knowledge nodes. Of course, the threshold value cannot be set too high to prevent subordinate knowledge nodes with high correlation with the target knowledge node from being filtered out. . When the degree of belonging of the subordinate knowledge node to the target knowledge node is less than the threshold, the subordinate knowledge node does not belong to the domain determined by the target knowledge node and needs to be eliminated. When the degree of belonging from the subordinate node to the target knowledge node is not less than a threshold, the subordinate knowledge node belongs to the domain determined by the target knowledge node. The remaining knowledge nodes are knowledge nodes that are determined from the preset knowledge items, and subordinate knowledge nodes that do not meet the requirements are removed. The non-conformity means that the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than Threshold.

Step S140, construct a knowledge graph according to the remaining knowledge nodes in the preset domain.

Among them, there is a one-to-one correspondence between knowledge nodes and knowledge items. According to the knowledge nodes, the corresponding knowledge items can be obtained, the corresponding knowledge items can be obtained through the knowledge nodes, and the value of the entity and the value of the entity attribute of the knowledge item can be obtained to complete the construction of the knowledge graph. .

Among them, the value of the entity and the value of the entity attribute of the knowledge item can be obtained from the encyclopedia database, and the value of the entity and the value of the entity attribute of the knowledge item can also be collected from the web page.

The above-mentioned knowledge graph construction method based on the degree of belonging of the knowledge node, by calculating the degree of belonging of the subordinate knowledge node to the target knowledge node, removes the subordinate knowledge node with low degree of relevance, ensuring the accuracy of the constructed knowledge graph, and passes The method of calculating the degree of belonging is used to streamline knowledge items, which improves the efficiency of knowledge graph construction.

In one of the embodiments, the step S110 includes: obtaining the association relationship between the knowledge node and the knowledge node according to the encyclopedia database; obtaining the target in the preset field according to the association relationship between the knowledge node and the knowledge node Knowledge node and subordinate knowledge node.

Among them, the knowledge node corresponds to the knowledge item. The knowledge node is obtained according to the encyclopedia database, and the knowledge item is first obtained from the encyclopedia database. The encyclopedia database is a database with open source data tables such as Wikipedia, and the knowledge in the data table is obtained through the open source data table interface The relationship between items and knowledge items, the encyclopedia database includes knowledge items in multiple fields. The association relationship between knowledge items is mapped to the association relationship between knowledge nodes. Among them, the composition of the knowledge item and the association relationship between the knowledge item is structured data, and the structured data is data with a fixed pattern and subordination relationship, such as data in a database. Semi-structured data: data with a basic fixed structure pattern, such as log files, XML documents, JSON documents, etc. Unstructured data: data without a fixed pattern, such as document data such as WORD, PDF, PPT, EXCEL, etc., pictures and videos in various formats.

For example, in Wikipedia, three tables of "page", "categoryLinks" and "redirect" are used to extract the relationship between classified knowledge entries and knowledge entries. Among them, the knowledge entries are stored in the "page" table. The association relationship between the knowledge items is stored in the "categoryLinks" and "redirect" tables.

Among them, the target knowledge node may be the largest knowledge concept in the preset field, and the subordinate knowledge node is a small category or sub-category under the target knowledge node. For example, if the target knowledge node is finance, it may exist with the finance. The subordinate knowledge nodes of subordinate relationship include currency, RMB, and U.S. dollar.

In one of the embodiments, as shown in FIG. 2, step S120 includes: step S121, obtaining intermediate knowledge nodes on the connection path of the target knowledge node to the subordinate knowledge node; step S122, calculating the target knowledge node and The number of knowledge nodes that the intermediate knowledge node leads to connections; step S123, calculate the reciprocal of the number of knowledge nodes that lead to connections, to obtain the link weights of the target knowledge node and the intermediate knowledge nodes; step S124, According to the link weight of the target knowledge node and the intermediate knowledge node, the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated.

Among them, the knowledge node and the knowledge node are connected by edges. For example, in Figure 3, T1, T2, C1, C2, C3, C4, and C5 are knowledge nodes, the arrows from T1 to C3, C4 are the edges of the knowledge node, and T2 to C1 The arrows of C4, C5 are the edges of the knowledge node, the arrows from C3 to C1, C2 are the edges of the knowledge node, and the arrows from C4 to C1 are the edges of the knowledge node; the intermediate knowledge node is on a connection path except for the target knowledge The node and other knowledge nodes of the subordinate knowledge node, as shown in Figure 3, a connection path T1-C3-C2, T1 is the target knowledge node, C2 is the subordinate knowledge node, then the intermediate knowledge node is not C3. Wherein, the number of knowledge nodes derived from the knowledge node is equal to the number of edges derived from the knowledge node. For example, as shown in Figure 3, two edges are derived from T1, then the connected knowledge node derived from T1 The number of is 2. On the connection path from the target knowledge node to the subordinate knowledge node, except for the subordinate knowledge node, each knowledge node counts the number of knowledge nodes that lead to the connection, and calculates the link weight of the knowledge node according to the number of knowledge nodes that lead to the connection. , The link weight is equal to the reciprocal of the number of knowledge nodes that lead to connections. For example, as shown in Figure 3, C3 leads to knowledge nodes C2 and C1, and the number of knowledge nodes that C3 leads to connections is 2, then the link weight of C3 is 1/2 .

This implementation eliminates sub-knowledge nodes with low relevance by calculating the degree of belonging of the subordinate knowledge node to the target knowledge node, without manual participation in the construction of the graph, and the degree of belonging threshold can be modified to change the recall rate and accuracy of the domain Spend. This application can specify different fields, and can quickly build a knowledge map of vertical fields.

In one of the embodiments, the step S134 includes: calculating the product of the link weights of the target knowledge node and the intermediate knowledge node of each connection path to obtain the weight of each connection path; the connection path Is the connection path from the subordinate knowledge node to the target knowledge node; calculates the sum of the weights of all the connection paths to obtain the degree of belonging of the subordinate knowledge node to the target knowledge node.

For example, as shown in Figure 3, the process of calculating the degree of belonging of C1 to T2 is as follows, T1, T2, C1, C2, C3, C4, and C5 are nodes, and the number of knowledge nodes connected to each knowledge node is equal to the number of knowledge nodes derived from it. The number of arrows, such as the number of T2 leading arrows is 3, the number of T1 leading arrows is 2, the number of C3 leading arrows is 2, the number of C4 leading arrows is 1, and the link weight of T2 is equal to 1/3 , The link weight of T1 is equal to 1/2, the link weight of C3 is equal to 1/2, and the link weight of C4 is equal to 1; there are two paths from T2 to C1, one of which must pass through C4, then the path T2, C4, C1 The weight of is 1/3×1=1/3, the weight of the path T2 and C1 is 1/3, then the degree of belonging of C1 to T2 is 1/3+1/3=2/3.

In one of the embodiments, the subordinate knowledge node is an end knowledge node; the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated according to the path length from the target knowledge node to the subordinate knowledge node , Including: calculating the degree of belonging of the end knowledge node to the target knowledge node according to the path length from the target knowledge node to the end knowledge node.

Among them, the end knowledge node is a knowledge node with no edges or zero edges. Since the end knowledge node is connected with the target knowledge node through multi-level knowledge nodes, the correlation between the end knowledge node and the target knowledge node is relatively small or not at all Associated, the end knowledge node needs to be pruned at this time to realize the optimization of the knowledge graph.

In one of the embodiments, the method for constructing a knowledge graph based on the degree of belonging of a knowledge node further includes: judging whether the degree of belonging of the subordinate knowledge node to the target knowledge node is less than a threshold. Wherein, the threshold value can be set as required, so as to expand and reduce the classification of the field. In one of the embodiments, the threshold can be set to 0.1.

In one of the embodiments, step S140 includes: obtaining remaining knowledge nodes in the preset domain; according to the remaining knowledge nodes, obtaining corresponding information boxes through crawler technology to construct knowledge items corresponding to the remaining knowledge nodes The value of the entity and the value of the entity attribute.

Wherein, according to the knowledge item corresponding to the remaining knowledge node, the corresponding information box is obtained in the Internet page by crawling technology. For example, the knowledge item is "currency", the entity and entity attributes are constructed for currency, and the value of the currency entity and the value of entity attributes are obtained through crawler technology. Specifically, the first step is to splice the request url such as: https://baike.baidu .com/item/currency, the second step uses xpath to obtain the information box data (the content in the solid line box 401 in Figure 4), and the third step saves the obtained information box data to the mongodb database.

In this embodiment, based on knowledge, structured data can be obtained by using crawler technology without parsing semi-structured data (news portal), which improves the efficiency of constructing a knowledge graph of vertical domains.

It should be understood that although the various steps in the flowcharts of FIGS. 1-2 are displayed in sequence as indicated by the arrows, these steps are not necessarily performed in sequence in the order indicated by the arrows. Unless specifically stated in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least part of the steps in Figure 1-2 can include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution of these steps or stages The sequence is not necessarily performed sequentially, but may be performed alternately or alternately with other steps or at least a part of the steps or stages in other steps.

In one embodiment, as shown in FIG. 5, a device for constructing a knowledge graph based on the degree of belonging of a knowledge node is provided, including: a knowledge node selection module 210, a degree of belonging calculation module 220, a node deletion module 230, and a knowledge graph constructing module 240.

The knowledge node selection module 210 is configured to obtain a target knowledge node and subordinate knowledge nodes of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain.

The degree of belonging calculation module 220 is configured to calculate the degree of belonging from the subordinate knowledge node to the target knowledge node according to the length of the path from the target knowledge node to the subordinate knowledge node; wherein, the path length is determined by the target knowledge node. The number of intermediate knowledge nodes connected from the knowledge node to the subordinate knowledge node is determined.

The node deletion module 230 is configured to delete the subordinate knowledge node when the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than a threshold value, and obtain the remaining knowledge nodes in the preset domain.

The knowledge graph construction module 240 is configured to construct a knowledge graph according to the remaining knowledge nodes in the preset domain.

In one of the embodiments, the knowledge node selection module 210 includes: an association relationship acquisition unit for acquiring an association relationship between a knowledge node and a knowledge node according to an encyclopedia database; a knowledge node acquisition unit for acquiring an association relationship between a knowledge node and a knowledge node according to the knowledge The relationship between the node and the knowledge node is to obtain the target knowledge node and the subordinate knowledge node in the preset field.

In one of the embodiments, the membership degree calculation module 220 includes: an intermediate knowledge node acquisition module for acquiring intermediate knowledge nodes on the connection path between the target knowledge node and the subordinate knowledge node; calculating the number of connected knowledge nodes A unit for calculating the number of knowledge nodes that are derived from the target knowledge node and the intermediate knowledge node; a link weight calculation unit for calculating the reciprocal of the number of knowledge nodes that are derived from the intermediate knowledge node to obtain the target The link weight of the knowledge node and the intermediate knowledge node; a membership degree calculation unit for calculating the membership degree of the subordinate knowledge node to the target knowledge node according to the link weight of the target knowledge node and the intermediate knowledge node .

In one of the embodiments, the belonging degree calculation unit includes: a weight calculation subunit for calculating the product of the link weights of the target knowledge node and the intermediate knowledge node of each connection path to obtain each of the The weight of the connection path; the connection path is the connection path from the subordinate knowledge node to the target knowledge node; the membership degree calculation subunit is used to calculate the sum of the weights of all the connection paths to obtain the subordinate knowledge node The degree of belonging to the target knowledge node.

In one of the embodiments, the subordinate knowledge node is an end knowledge node; the membership degree calculation module 220 is further configured to calculate the end knowledge node according to the path length from the target knowledge node to the end knowledge node The degree of belonging to the target knowledge node.

In one of the embodiments, the knowledge graph construction device further includes: a judging module, configured to judge whether the degree of belonging of the subordinate knowledge node to the target knowledge node is less than a threshold; wherein the threshold is 0.1.

In one of the embodiments, the knowledge graph construction module 240 includes: a remaining knowledge node acquiring unit, configured to acquire remaining knowledge nodes in the preset domain; a value construction unit, configured to pass through the remaining knowledge nodes The crawler technology obtains the corresponding information box to construct the value of the entity and the value of the entity attribute of the knowledge item corresponding to the remaining knowledge node.

For the specific definition of the knowledge graph construction device based on the degree of belonging of a knowledge node, please refer to the above restriction on the method of constructing a knowledge graph based on the degree of belonging of a knowledge node, which will not be repeated here. Each module in the apparatus for constructing a knowledge graph based on the degree of belonging of a knowledge node can be implemented in whole or in part by software, hardware, and a combination thereof. The above-mentioned modules may be embedded in the form of hardware or independent of the processor in the computer equipment, or may be stored in the memory of the computer equipment in the form of software, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. 6. The computer equipment includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device is used to store the data of the knowledge item and the association relationship between the knowledge item. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for constructing a knowledge graph based on the degree of belonging of the knowledge node is realized.

Those skilled in the art can understand that the structure shown in FIG. 6 is only a block diagram of part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may Including more or fewer parts than shown in the figure, or combining some parts, or having a different arrangement of parts.

In one embodiment, a computer device is provided, including a memory and a processor, and a computer program is stored in the memory. When the processor executes the computer program, the method for constructing a knowledge graph based on the degree of belonging of a knowledge node is implemented, which is based on the knowledge node The method for constructing the knowledge graph of the degree of belonging includes the following steps: obtaining a target knowledge node and subordinate knowledge nodes of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain; according to the target The length of the path from the knowledge node to the subordinate knowledge node is calculated, and the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated; wherein, the path length is determined by the connection between the target knowledge node and the subordinate knowledge node The number of intermediate knowledge nodes is determined; when the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than a threshold, the subordinate knowledge node is deleted, and the remaining knowledge nodes in the preset domain are obtained; according to the The remaining knowledge nodes in the preset field construct a knowledge graph.

In an embodiment, the processor further implements the following steps when executing the computer program: judging whether the degree of belonging of the subordinate knowledge node to the target knowledge node is less than a threshold; wherein the threshold is 0.1.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method for constructing a knowledge graph based on the degree of belonging of a knowledge node is realized. The knowledge graph construction method includes the following steps: acquiring a target knowledge node and subordinate knowledge nodes of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain; The path length of the subordinate knowledge node calculates the degree of belonging of the subordinate knowledge node to the target knowledge node; wherein the path length is from the target knowledge node to the intermediate knowledge node connected to the subordinate knowledge node The number of is determined; when the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than the threshold, the subordinate knowledge node is deleted, and the remaining knowledge nodes in the preset domain are obtained; according to the preset domain The remaining knowledge nodes inside construct a knowledge graph.

In one embodiment, when the computer program is executed by the processor, the following steps are further implemented: judging whether the degree of belonging of the subordinate knowledge node to the target knowledge node is less than a threshold; wherein the threshold is 0.1.

Optionally, the storage medium involved in this application, such as a computer-readable storage medium, may be non-volatile or volatile.

A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage. In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, or optical memory, etc. Volatile memory may include Random Access Memory (RAM) or external cache memory. As an illustration and not a limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM).

The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be It is considered as the range described in this specification.

The above-mentioned embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it should not be understood as a limitation on the scope of the invention patent. It should be pointed out that for those of ordinary skill in the art, without departing from the concept of this application, several modifications and improvements can be made, and these all fall within the protection scope of this application. Therefore, the scope of protection of the patent of this application shall be subject to the appended claims.

Claims

A method for constructing a knowledge graph based on the degree of belonging of a knowledge node, wherein the method includes:

Acquiring a target knowledge node and a subordinate knowledge node of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain;

According to the path length from the target knowledge node to the subordinate knowledge node, the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated; wherein the path length is from the target knowledge node to the subordinate knowledge node Determine the number of intermediate knowledge nodes connected in;

When the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than a threshold, delete the subordinate knowledge node, and obtain the remaining knowledge nodes in the preset domain;

Construct a knowledge graph according to the remaining knowledge nodes in the preset domain.
The method according to claim 1, wherein said acquiring the target knowledge node and the subordinate knowledge nodes of the target knowledge node comprises:

According to the encyclopedia database, obtain the relationship between the knowledge node and the knowledge node;

According to the association relationship between the knowledge node and the knowledge node, the target knowledge node and the subordinate knowledge node in the preset field are obtained.
The method according to claim 1, wherein the calculating the degree of belonging of the subordinate knowledge node to the target knowledge node according to the length of the path from the target knowledge node to the subordinate knowledge node comprises:

Acquiring an intermediate knowledge node on a connection path from the target knowledge node to the subordinate knowledge node;

Calculating the number of knowledge nodes that are connected by the target knowledge node and the intermediate knowledge node;

Calculate the reciprocal of the number of knowledge nodes that lead to connections, and obtain the link weights of the target knowledge nodes and the intermediate knowledge nodes;

According to the link weight of the target knowledge node and the intermediate knowledge node, the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated.
The method according to claim 3, wherein the calculating the degree of belonging of the subordinate knowledge node to the target knowledge node according to the link weight of the target knowledge node and the intermediate knowledge node comprises:

Calculate the product of the link weights of the target knowledge node and the intermediate knowledge node for each connection path to obtain the weight of each connection path; the connection path is the distance from the subordinate knowledge node to the target knowledge node Connection path

Calculate the sum of the weights of all the connection paths to obtain the degree of belonging of the subordinate knowledge node to the target knowledge node.
The method according to claim 1, wherein the subordinate knowledge node is an end knowledge node;

The calculating the degree of belonging of the subordinate knowledge node to the target knowledge node according to the path length from the target knowledge node to the subordinate knowledge node includes:

According to the path length from the target knowledge node to the end knowledge node, the degree of belonging of the end knowledge node to the target knowledge node is calculated.
The method according to claim 1, further comprising: determining whether the degree of belonging of the subordinate knowledge node to the target knowledge node is less than a threshold; wherein the threshold is 0.1.
The method according to claim 1, wherein the constructing a knowledge graph based on the remaining knowledge nodes in the preset domain comprises:

Acquiring remaining knowledge nodes in the preset domain;

According to the remaining knowledge node, the corresponding information box is obtained through crawler technology to construct the value of the entity and the value of the entity attribute of the knowledge item corresponding to the remaining knowledge node.
A device for constructing a knowledge graph based on the degree of belonging of a knowledge node, wherein the device includes:

The knowledge node selection module is used to obtain a target knowledge node and subordinate knowledge nodes of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain;

The degree of belonging calculation module is used to calculate the degree of belonging of the subordinate knowledge node to the target knowledge node according to the length of the path from the target knowledge node to the subordinate knowledge node; wherein, the path length is determined by the target knowledge node. Determining the number of intermediate knowledge nodes connected from the node to the subordinate knowledge node;

A node deletion module, configured to delete the subordinate knowledge node when the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than a threshold value, and obtain the remaining knowledge nodes in the preset domain;

The knowledge graph construction module is used to construct a knowledge graph according to the remaining knowledge nodes in the preset domain.
A computer device includes a memory and a processor, the memory stores a computer program, wherein the processor implements a method for constructing a knowledge graph based on the degree of belonging of a knowledge node when the processor executes the computer program, and the method for constructing a knowledge graph based on the degree of belonging of a knowledge node is The knowledge graph construction method includes the following steps:

Acquiring a target knowledge node and a subordinate knowledge node of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain;

According to the path length from the target knowledge node to the subordinate knowledge node, the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated; wherein the path length is from the target knowledge node to the subordinate knowledge node Determine the number of intermediate knowledge nodes connected in;

When the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than a threshold, delete the subordinate knowledge node, and obtain the remaining knowledge nodes in the preset domain;

Construct a knowledge graph according to the remaining knowledge nodes in the preset domain.
The computer device according to claim 9, wherein the execution of the acquisition target knowledge node and the subordinate knowledge node of the target knowledge node comprises:

According to the encyclopedia database, obtain the relationship between the knowledge node and the knowledge node;

According to the association relationship between the knowledge node and the knowledge node, the target knowledge node and the subordinate knowledge node in the preset field are obtained.
8. The computer device according to claim 9, wherein the calculation of the degree of belonging of the subordinate knowledge node to the target knowledge node according to the path length from the target knowledge node to the subordinate knowledge node comprises:

Acquiring an intermediate knowledge node on a connection path from the target knowledge node to the subordinate knowledge node;

Calculating the number of knowledge nodes that are connected by the target knowledge node and the intermediate knowledge node;

Calculate the reciprocal of the number of knowledge nodes that lead to connections, and obtain the link weights of the target knowledge nodes and the intermediate knowledge nodes;

According to the link weight of the target knowledge node and the intermediate knowledge node, the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated.
The computer device according to claim 9, wherein the subordinate knowledge node is an end knowledge node;

The execution of calculating the degree of belonging of the subordinate knowledge node to the target knowledge node according to the path length from the target knowledge node to the subordinate knowledge node includes:

According to the path length from the target knowledge node to the end knowledge node, the degree of belonging of the end knowledge node to the target knowledge node is calculated.
The computer device according to claim 9, wherein, when the processor executes the method for constructing a knowledge graph based on the degree of belonging of a knowledge node, the method further comprises:

Determine whether the degree of belonging of the subordinate knowledge node to the target knowledge node is less than a threshold; wherein the threshold is 0.1.
The computer device according to claim 9, wherein the execution of the construction of the knowledge graph based on the remaining knowledge nodes in the preset domain comprises:

Acquiring remaining knowledge nodes in the preset domain;

According to the remaining knowledge node, the corresponding information box is obtained through crawler technology to construct the value of the entity and the value of the entity attribute of the knowledge item corresponding to the remaining knowledge node.
A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, a method for constructing a knowledge graph based on the degree of belonging of a knowledge node is implemented, and the method for constructing a knowledge graph based on the degree of belonging of a knowledge node includes The following steps:

Acquiring a target knowledge node and a subordinate knowledge node of the target knowledge node; the target knowledge node and the subordinate knowledge node are knowledge nodes in a preset domain;

According to the path length from the target knowledge node to the subordinate knowledge node, the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated; wherein the path length is from the target knowledge node to the subordinate knowledge node Determine the number of intermediate knowledge nodes connected in;

When the degree of belonging of the subordinate knowledge node to the target knowledge node is lower than a threshold, delete the subordinate knowledge node, and obtain the remaining knowledge nodes in the preset domain;

Construct a knowledge graph according to the remaining knowledge nodes in the preset domain.
The computer-readable storage medium according to claim 15, wherein the execution of the acquisition target knowledge node and the subordinate knowledge node of the target knowledge node comprises:

According to the encyclopedia database, obtain the relationship between the knowledge node and the knowledge node;

According to the association relationship between the knowledge node and the knowledge node, the target knowledge node and the subordinate knowledge node in the preset field are obtained.
The computer-readable storage medium according to claim 15, wherein when performing the calculation of the degree of belonging of the subordinate knowledge node to the target knowledge node based on the path length from the target knowledge node to the subordinate knowledge node ,include:

Acquiring an intermediate knowledge node on a connection path from the target knowledge node to the subordinate knowledge node;

Calculating the number of knowledge nodes that are connected by the target knowledge node and the intermediate knowledge node;

Calculate the reciprocal of the number of knowledge nodes that lead to connections, and obtain the link weights of the target knowledge nodes and the intermediate knowledge nodes;

According to the link weight of the target knowledge node and the intermediate knowledge node, the degree of belonging of the subordinate knowledge node to the target knowledge node is calculated.
The computer-readable storage medium according to claim 15, wherein the subordinate knowledge node is an end knowledge node;

The execution of calculating the degree of belonging of the subordinate knowledge node to the target knowledge node according to the path length from the target knowledge node to the subordinate knowledge node includes:

According to the path length from the target knowledge node to the end knowledge node, the degree of belonging of the end knowledge node to the target knowledge node is calculated.
The computer-readable storage medium according to claim 15, wherein, when the computer program is executed by a processor to implement the method for constructing a knowledge graph based on the degree of belonging of a knowledge node, the method further comprises:

Determine whether the degree of belonging of the subordinate knowledge node to the target knowledge node is less than a threshold; wherein the threshold is 0.1.
The computer-readable storage medium according to claim 15, wherein the execution of the construction of the knowledge graph based on the remaining knowledge nodes in the preset field comprises:

Acquiring remaining knowledge nodes in the preset domain;

According to the remaining knowledge node, the corresponding information box is obtained through crawler technology to construct the value of the entity and the value of the entity attribute of the knowledge item corresponding to the remaining knowledge node.