A method and device for automatically associating knowledge in a knowledge base

Abstract

The present invention discloses a method and device for automatically associating knowledge in a knowledge base, which optimizes the parameters of a relationship analysis model by using historical data of multiple target dimensions of multiple pieces of knowledge in the knowledge base as training samples to obtain a trained relationship analysis model, and calculates the association coefficient of each piece of knowledge in the knowledge base through the trained relationship analysis model, and then establishes the association relationship between the knowledge through the association coefficient. It can be seen that the degree of association between each piece of knowledge in the knowledge base can be analyzed through the pre-trained relationship analysis model, so that when the user searches for related knowledge, the user can be provided with knowledge that better meets the user's expectations.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to a method and device for automatically associating knowledge in a knowledge base. Background Art

[0002] The knowledge base is a massive knowledge management tool platform, and its main function is to store and manage knowledge and provide knowledge retrieval capabilities. Currently, the knowledge in the knowledge base is independent of each other, and when users search, they cannot discover their potential needs.

[0003] In the prior art, in order to solve the above problems, the association between knowledge is manually established. This method has low processing efficiency and is relatively subjective, making it difficult to discover the implicit association information between knowledge. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a method and device for automatically associating knowledge in a knowledge base. Through a pre-trained relationship analysis model, the degree of association between each piece of knowledge in the knowledge base can be analyzed, which not only achieves the purpose of intelligent knowledge association, but also the established association relationship can reflect the degree of association between knowledge. When a user searches for related knowledge, the user can be provided with knowledge that better meets the user's expectations. Moreover, when associating knowledge, information of multiple dimensions is considered, and the possible association relationship between knowledge is considered more comprehensively, so that the association relationship between knowledge established is more accurate.

[0005] The embodiment of the present invention discloses a method for automatically associating knowledge in a knowledge base, comprising:

[0006] Acquire information of multiple target dimensions of each piece of knowledge in the knowledge base; the information of the multiple target dimensions at least includes: attribute information of the knowledge and / or content information of the knowledge;

[0007] Extracting a first target feature from information of multiple target dimensions of each piece of knowledge;

[0008] Inputting the first target feature into a preset relationship analysis model to obtain the association coefficient of each piece of knowledge; wherein the relationship analysis model is obtained by optimizing the parameters of the relationship analysis model using historical data of multiple target dimensions of knowledge in the knowledge base as training samples;

[0009] Establish the association relationship between each piece of knowledge in the knowledge base according to the association coefficient of the knowledge.

[0010] Optionally, the attribute knowledge includes:

[0011] Knowledge title, creator, creation time, directory, tags, and knowledge source.

[0012] Optionally, the content information of the knowledge includes:

[0013] Knowledge keyword similarity information and text similarity information.

[0014] Optionally, the training process of the association model includes:

[0015] Acquire a second target feature of historical data of multiple target dimensions of multiple pieces of knowledge;

[0016] The second features of the historical data of multiple target dimensions of the multiple pieces of knowledge are input as training samples into the association model to be trained, and the association model to be trained is trained with the purpose of making the association coefficient output by the association model to be trained for the training samples approach the association coefficient of the pre-set multiple pieces of knowledge.

[0017] Optionally, the second target feature of the historical data of multiple target dimensions of multiple pieces of knowledge is obtained, including:

[0018] Obtain historical data of preset multiple dimensions from the user's account information;

[0019] Obtaining a second target feature from the historical data of each dimension;

[0020] Determining a weight for each second target feature of the history of the dimension;

[0021] According to the weight of the second target feature of the historical data of each dimension, the second target feature of the target dimension is screened out from the multiple dimensions.

[0022] The embodiment of the present invention discloses an automatic association device for knowledge in a knowledge base, comprising:

[0023] An acquisition unit, configured to acquire information of multiple target dimensions of each piece of knowledge in the knowledge base; the information of the multiple target dimensions at least includes: attribute information of the knowledge and / or content information of the knowledge;

[0024] A feature extraction unit, used to extract a first target feature from the information of multiple target dimensions of each piece of knowledge;

[0025] A correlation coefficient calculation unit, used for inputting the first target feature into a preset relationship analysis model to obtain a correlation coefficient of each piece of knowledge; wherein the relationship analysis model is obtained by optimizing the parameters of the relationship analysis model by using historical data of multiple target dimensions of knowledge in the knowledge base as training samples;

[0026] The relationship building unit is used to establish the relationship between each piece of knowledge in the knowledge base according to the association coefficient of the knowledge.

[0027] Optionally, the attribute knowledge includes:

[0028] Knowledge title, creator, creation time, directory, tags, and knowledge source.

[0029] Optionally, the content information of the knowledge includes:

[0030] Knowledge keyword similarity information and text similarity information.

[0031] Optionally, also include:

[0032] Association model training unit:

[0033] Acquire a second target feature of historical data of multiple target dimensions of multiple pieces of knowledge;

[0034] The second features of the historical data of multiple target dimensions of the multiple pieces of knowledge are input as training samples into the association model to be trained, and the association model to be trained is trained with the purpose of making the association coefficient output by the association model to be trained for the training samples approach the association coefficient of the pre-set multiple pieces of knowledge.

[0035] The embodiment of the present invention further discloses an electronic device, including:

[0036] Memory and processor;

[0037] The memory is used to store programs, and the processor is used to at least execute the following automatic association method of knowledge in a knowledge base:

[0038] Acquire information of multiple target dimensions of each piece of knowledge in the knowledge base; the information of the multiple target dimensions at least includes: attribute information of the knowledge and / or content information of the knowledge;

[0039] Extracting a first target feature from information of multiple target dimensions of each piece of knowledge;

[0040] Inputting the first target feature into a preset relationship analysis model to obtain the association coefficient of each piece of knowledge; wherein the relationship analysis model is obtained by optimizing the parameters of the relationship analysis model using historical data of multiple target dimensions of knowledge in the knowledge base as training samples;

[0041] Establish the association relationship between each piece of knowledge in the knowledge base according to the association coefficient of the knowledge.

[0042] The embodiment of the present invention discloses a method and device for automatically associating knowledge in a knowledge base, which optimizes the parameters of a relationship analysis model by using historical data of multiple target dimensions of multiple pieces of knowledge in the knowledge base as training samples to obtain a trained relationship analysis model, and calculates the correlation coefficient of each piece of knowledge in the knowledge base through the trained relationship analysis model, and then establishes the correlation relationship between the knowledge through the correlation coefficient. It can be seen that the degree of correlation between the various pieces of knowledge in the knowledge base can be analyzed through the pre-trained relationship analysis model, which not only achieves the purpose of intelligent knowledge association, but also the established correlation relationship can reflect the degree of correlation between the knowledge. When the user searches for related knowledge, the user can be provided with knowledge that better meets the user's expectations. In addition, when performing knowledge association, information of multiple dimensions is considered, and the possible correlation relationship between knowledge is more comprehensively considered, so that the established correlation relationship between knowledge is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 A schematic diagram of a process of automatically associating knowledge in a knowledge base provided by an embodiment of the present invention is shown;

[0045] Figure 2 A schematic diagram of a flow chart of a method for training a relationship analysis model provided by an embodiment of the present invention;

[0046] Figure 3 A schematic diagram showing the structure of a device for automatically associating knowledge in a knowledge base provided by an embodiment of the present invention is shown;

[0047] Figure 4 A schematic structural diagram of an electronic device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] refer to Figure 1, shows a schematic flow chart of a method for automatically associating knowledge in a knowledge base provided by an embodiment of the present invention. In this embodiment, the method includes:

[0050] S101: Acquire information of multiple target dimensions of each piece of knowledge in a knowledge base; the information of the multiple target dimensions at least includes: attribute information of the knowledge and / or content information of the knowledge;

[0051] In this embodiment, the attribute information of knowledge may include multiple types. In this embodiment, the attribute information selected is the attribute information that affects the correlation between knowledge, or the attribute information that has a greater impact on the correlation between knowledge. Among them, the attribute information selected in this embodiment may include: knowledge title, creator, creation time, label, knowledge source, etc.

[0052] The content information of knowledge may include various types, for example, the content of the knowledge itself.

[0053] Or, in order to further improve the accuracy of knowledge association, the content information of the knowledge can adopt keyword similarity and text similarity.

[0054] There are many methods for obtaining keyword similarity, which are not limited in this embodiment. Preferably, the following method can be used to obtain keyword similarity:

[0055] First, stop words are removed, and then the knowledge content is segmented to form several keywords. The proportion of the same keywords between any knowledge is calculated. The proportion of the same keywords can be calculated by the following formula 1):

[0056] 1) Keyword ratio = number of identical keywords / total number of keywords in the knowledge;

[0057] For example, to calculate the keyword similarity between knowledge A and knowledge B, the keyword similarity between knowledge A and knowledge B = the number of identical keywords between knowledge A and knowledge B / the number of keywords in knowledge A).

[0058] In this embodiment, text similarity can be understood as the similarity of the content between two pieces of knowledge, wherein the text similarity between pieces of knowledge can be obtained through a variety of methods, which are not limited in this embodiment.

[0059] In this embodiment, the knowledge stored in the knowledge base includes information of multiple dimensions, among which some dimensions of information have a greater impact on the association of knowledge, and some dimensions of information may have a smaller impact on the association between knowledge. In this embodiment, information of multiple dimensions can be screened out by feature engineering. Preferably, the following method can be used to obtain the information:

[0060] Obtaining historical information of preset multiple dimensions from the user's account information;

[0061] Obtaining a second target feature from the historical information of each dimension;

[0062] Determine the weight of the second target feature of the historical information of each dimension;

[0063] According to the weight of the second target feature of the historical information of each dimension, a target dimension is screened out from the multiple dimensions.

[0064] S102: extracting a first target feature from information of multiple target dimensions of each piece of knowledge;

[0065] In this embodiment, there are multiple methods for extracting the first target feature from information of multiple target dimensions of knowledge, which are not limited in this embodiment.

[0066] Preferably, a feature engineering model may be used for feature extraction. The feature engineering model may include multiple types, which are not limited in this embodiment. For example, a principal component analysis model, a linear discriminant analysis model, etc. may be used.

[0067] S103: Inputting the first target feature into a preset relationship analysis model to obtain a correlation coefficient of each piece of knowledge;

[0068] The relationship analysis model is obtained by optimizing and training the parameters of the relationship analysis model using historical data of multiple target dimensions of knowledge in the knowledge base as training samples.

[0069] In this embodiment, the relationship analysis model can be any model that can perform clustering. Before executing S103, the relationship analysis model is pre-trained. Specifically, the parameters in the relationship analysis model are optimized through training. The specific training method will be introduced in detail below and will not be repeated in this embodiment.

[0070] S104: Establishing association relationships between various pieces of knowledge in the knowledge base according to the association coefficient of the knowledge.

[0071] In this embodiment, the correlation coefficient between different knowledge can be understood as the degree of correlation between different knowledge, wherein the larger the correlation coefficient is, the higher the degree of correlation between the knowledge is considered to be.

[0072] In this embodiment, the parameters of the relationship analysis model are optimized by using the historical data of multiple target dimensions of multiple pieces of knowledge in the knowledge base as training samples to obtain a trained relationship analysis model, and the correlation coefficient of each piece of knowledge in the knowledge base is calculated by the trained relationship analysis model, and then the correlation relationship between the knowledge is established by the correlation coefficient. It can be seen that the degree of correlation between the various pieces of knowledge in the knowledge base can be analyzed through the pre-trained relationship analysis model, so that when the user searches for related knowledge, the user can be provided with knowledge that better meets the user's expectations.

[0073] Furthermore, when new knowledge is received in the knowledge base, the correlation coefficient between the knowledge and other knowledge can be determined through the pre-trained relationship analysis model, specifically, including:

[0074] Acquire information of multiple target dimensions of target knowledge, and extract a third target feature from the information of multiple target dimensions of target knowledge;

[0075] The third target feature is input into the relationship analysis model to obtain the correlation coefficient between the knowledge and other knowledge in the knowledge base.

[0076] Among them, the relational analysis model pre-clustered the existing knowledge in the knowledge base, so the relational analysis model stores the clustering results of the existing knowledge in the knowledge base, and the correlation coefficient between the newly added target knowledge and the existing knowledge can be determined through the clustering results of the existing knowledge in the knowledge base.

[0077] refer to Figure 2 , the relationship analysis model can also be trained by the following methods, including:

[0078] S201: Acquire a second target feature of historical data of multiple target dimensions of multiple pieces of knowledge in a knowledge base;

[0079] In this embodiment, the knowledge stored in the knowledge base includes information of multiple dimensions, among which some dimensions of information have a greater impact on the association of knowledge, and some dimensions of information may have a smaller impact on the association between knowledge. In this embodiment, information of multiple dimensions can be screened out by feature engineering. Preferably, the following method can be used to obtain the information:

[0080] Obtaining historical information of preset multiple dimensions from the user's account information;

[0081] Obtaining a second target feature from the historical information of each dimension;

[0082] Determine the weight of the second target feature of the historical information of each dimension;

[0083] According to the weight of the second target feature of the historical information of each dimension, the second target features of multiple target dimensions are screened out from the multiple dimensions.

[0084] S202: Input the second target features of the historical data of multiple target dimensions of multiple pieces of knowledge in the knowledge base as training samples into the trained association model, and train the association model to be trained with the aim of making the association coefficient output by the relationship analysis model to be trained for the training samples approach the association coefficient of the pre-set knowledge.

[0085] In this embodiment, the pre-set knowledge correlation coefficient can be set based on existing knowledge or experience.

[0086] In this embodiment, the pre-trained relationship analysis model can be used to analyze the degree of association between various pieces of knowledge in the knowledge base, so that when the user searches for related knowledge, the user can be provided with knowledge that better meets the user's expectations.

[0087] Alternatively, the relationship analysis model can also be trained by other methods, such as training by an unsupervised machine learning model. The process may include:

[0088] S301: extracting a second target feature from historical data of multiple target dimensions of multiple pieces of knowledge in a knowledge base;

[0089] In this embodiment, the knowledge stored in the knowledge base includes information of multiple dimensions, among which some dimensions of information have a greater impact on the association of knowledge, and some dimensions of information may have a smaller impact on the association between knowledge. In this embodiment, information of multiple dimensions can be screened out by feature engineering. Preferably, the following method can be used to obtain the information:

[0090] Obtaining historical information of preset multiple dimensions from the user's account information;

[0091] Obtaining a second target feature from the historical information of each dimension;

[0092] Determine the weight of the second target feature of the historical information of each dimension;

[0093] According to the weight of the second target feature of the historical information of each dimension, the second target features of multiple target dimensions are screened out from the multiple dimensions.

[0094] S302: Input the second target feature into a relationship analysis model, and optimize the relationship analysis model by adjusting the target.

[0095] In this embodiment, the target optimization function may be preset, for example, a loss function may be constructed by the similarity of each piece of knowledge output, or a loss function may be constructed by the similarity of each piece of knowledge feature.

[0096] In this embodiment, the pre-trained relationship analysis model can be used to analyze the degree of association between various pieces of knowledge in the knowledge base, so that when the user searches for related knowledge, the user can be provided with knowledge that better meets the user's expectations.

[0097] refer to Figure 3 , shows a schematic diagram of the structure of a device for automatically associating knowledge in a knowledge base provided by an embodiment of the present invention. In this embodiment, the device includes:

[0098] The acquisition unit 401 is used to acquire information of multiple target dimensions of each piece of knowledge in the knowledge base; the information of the multiple target dimensions at least includes: attribute information of the knowledge and / or content information of the knowledge;

[0099] A feature extraction unit 402, configured to extract a first target feature from information of multiple target dimensions of each piece of knowledge;

[0100] The association coefficient calculation unit 403 is used to input the first target feature into a preset relationship analysis model to obtain the association coefficient of each piece of knowledge; wherein the relationship analysis model is obtained by optimizing the parameters of the relationship analysis model by using historical data of multiple target dimensions of knowledge in the knowledge base as training samples;

[0101] The relationship building unit 404 is used to build the relationship between each piece of knowledge in the knowledge base according to the association coefficient of the knowledge.

[0102] Optionally, the attribute knowledge includes:

[0103] Knowledge title, creator, creation time, directory, tags, and knowledge source.

[0104] Optionally, the content information of the knowledge includes:

[0105] Knowledge keyword similarity information and text similarity information.

[0106] Optionally, also include:

[0107] Association model training unit:

[0108] Acquire a second target feature of historical data of multiple target dimensions of multiple pieces of knowledge;

[0109] The second features of the historical data of multiple target dimensions of the multiple pieces of knowledge are input as training samples into the association model to be trained, and the association model to be trained is trained with the purpose of making the association coefficient output by the association model to be trained for the training samples approach the association coefficient of the pre-set multiple pieces of knowledge.

[0110] Optionally, the association model training unit further includes:

[0111] Obtain historical data of preset multiple dimensions from the user's account information;

[0112] Obtaining a second target feature from the historical data of each dimension;

[0113] Determining a weight for each second target feature of the history of the dimension;

[0114] According to the weight of the second target feature of the historical data of each dimension, the second target feature of the target dimension is screened out from the multiple dimensions.

[0115] The device of this embodiment optimizes the parameters of the relationship analysis model by using the historical data of multiple target dimensions of multiple pieces of knowledge in the knowledge base as training samples to obtain a trained relationship analysis model, and calculates the correlation coefficient of each piece of knowledge in the knowledge base through the trained relationship analysis model, and then establishes the correlation relationship between the knowledge through the correlation coefficient. It can be seen that the degree of correlation between the various pieces of knowledge in the knowledge base can be analyzed through the pre-trained relationship analysis model, so that when the user searches for related knowledge, the user can be provided with knowledge that better meets the user's expectations.

[0116] refer to Figure 4 , shows a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. In this embodiment, the electronic device includes:

[0117] Memory 401 and processor 402;

[0118] The memory is used to store programs, and the processor is used to execute the following method for automatically associating knowledge in a knowledge base:

[0119] Acquire information of multiple target dimensions of each piece of knowledge in the knowledge base; the information of the multiple target dimensions at least includes: attribute information of the knowledge and / or content information of the knowledge;

[0120] Extracting a first target feature from information of multiple target dimensions of each piece of knowledge;

[0121] Inputting the first target feature into a preset relationship analysis model to obtain the association coefficient of each piece of knowledge; wherein the relationship analysis model is obtained by optimizing the parameters of the relationship analysis model using historical data of multiple target dimensions of knowledge in the knowledge base as training samples;

[0122] Establish the association relationship between each piece of knowledge in the knowledge base according to the association coefficient of the knowledge.

[0123] Optionally, the attribute knowledge includes:

[0124] Knowledge title, creator, creation time, directory, tags, and knowledge source.

[0125] Optionally, the content information of the knowledge includes:

[0126] Knowledge keyword similarity information and text similarity information.

[0127] Optionally, the training process of the association model includes:

[0128] Acquire a second target feature of historical data of multiple target dimensions of multiple pieces of knowledge;

[0129] The second features of the historical data of multiple target dimensions of the multiple pieces of knowledge are input as training samples into the association model to be trained, and the association model to be trained is trained with the purpose of making the association coefficient output by the association model to be trained for the training samples approach the association coefficient of the pre-set multiple pieces of knowledge.

[0130] Optionally, the second target feature of the historical data of multiple target dimensions of multiple pieces of knowledge is obtained, including:

[0131] Obtain historical data of preset multiple dimensions from the user's account information;

[0132] Obtaining a second target feature from the historical data of each dimension;

[0133] Determining a weight for each second target feature of the history of the dimension;

[0134] According to the weight of the second target feature of the historical data of each dimension, the second target feature of the target dimension is screened out from the multiple dimensions.

[0135] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0136] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0137] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0138] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0139] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0140] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0141] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0142] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for automatically associating knowledge in a knowledge base, characterized in that: include: Obtain information on multiple target dimensions for each piece of knowledge in the knowledge base; The information of the multiple target dimensions includes at least: attribute information of knowledge and / or content information of knowledge; Extracting a first target feature from information of multiple target dimensions of each piece of knowledge; Inputting the first target feature into a preset relationship analysis model to obtain the association coefficient of each piece of knowledge; wherein the relationship analysis model is obtained by optimizing the parameters of the relationship analysis model using historical data of multiple target dimensions of knowledge in the knowledge base as training samples; Establish the association relationship between each piece of knowledge in the knowledge base according to the association coefficient of the knowledge; The training process of the relationship analysis model includes: Obtain historical data of preset multiple dimensions from the user's account information; Obtaining a second target feature from the historical data of each dimension; Determining a weight for each second target feature of the history of the dimension; According to the weight of the second target feature of the historical data of each dimension, the second target feature of the target dimension is screened out from the multiple dimensions; The second features of the historical data of multiple target dimensions of multiple pieces of knowledge are input as training samples into the relationship analysis model to be trained, and the relationship analysis model to be trained is trained with the purpose of making the correlation coefficient output by the relationship analysis model to be trained for the training samples approach the correlation coefficient of the pre-set multiple pieces of knowledge.

2. The method according to claim 1, characterized in that: The attribute information of the knowledge includes: Knowledge title, creator, creation time, directory, tags, and knowledge source.

3. The method according to claim 1, characterized in that The content information of the knowledge includes: Knowledge keyword similarity information and text similarity information.

4. An automatic association device for knowledge in a knowledge base, characterized in that: include: An acquisition unit, used for acquiring information of multiple target dimensions of each piece of knowledge in the knowledge base; The information of the multiple target dimensions includes at least: attribute information of knowledge and / or content information of knowledge; A feature extraction unit, used to extract a first target feature from the information of multiple target dimensions of each piece of knowledge; A correlation coefficient calculation unit, used for inputting the first target feature into a preset relationship analysis model to obtain a correlation coefficient of each piece of knowledge; wherein the relationship analysis model is obtained by optimizing the parameters of the relationship analysis model by using historical data of multiple target dimensions of knowledge in the knowledge base as training samples; A relationship building unit is used to establish the relationship between each piece of knowledge in the knowledge base according to the knowledge association coefficient; Also includes: Relational analysis model training unit, used for: Obtain historical data of preset multiple dimensions from the user's account information; Obtaining a second target feature from the historical data of each dimension; Determining a weight for each second target feature of the history of the dimension; According to the weight of the second target feature of the historical data of each dimension, the second target feature of the target dimension is screened out from the multiple dimensions; The second features of the historical data of multiple target dimensions of multiple pieces of knowledge are input as training samples into the relationship analysis model to be trained, and the relationship analysis model to be trained is trained with the purpose of making the correlation coefficient output by the relationship analysis model to be trained for the training samples approach the correlation coefficient of the pre-set multiple pieces of knowledge.

5. The device according to claim 4, characterized in that The attribute information of the knowledge includes: Knowledge title, creator, creation time, directory, tags, and knowledge source.

6. The device according to claim 4, characterized in that The content information of the knowledge includes: Knowledge keyword similarity information and text similarity information.

7. An electronic device, characterized in that: include: Memory and processor; The memory is used to store programs, and the processor is used to at least execute the following automatic association method of knowledge in a knowledge base: Obtain information on multiple target dimensions for each piece of knowledge in the knowledge base; The information of the multiple target dimensions includes at least: attribute information of knowledge and / or content information of knowledge; Extracting a first target feature from information of multiple target dimensions of each piece of knowledge; Inputting the first target feature into a preset relationship analysis model to obtain the association coefficient of each piece of knowledge; wherein the relationship analysis model is obtained by optimizing the parameters of the relationship analysis model using historical data of multiple target dimensions of knowledge in the knowledge base as training samples; Establish the association relationship between each piece of knowledge in the knowledge base according to the association coefficient of the knowledge; The training process of the relationship analysis model includes: Obtain historical data of preset multiple dimensions from the user's account information; Obtaining a second target feature from the historical data of each dimension; Determining a weight for each second target feature of the history of the dimension; According to the weight of the second target feature of the historical data of each dimension, the second target feature of the target dimension is screened out from the multiple dimensions; The second features of the historical data of multiple target dimensions of multiple pieces of knowledge are input as training samples into the relationship analysis model to be trained, and the relationship analysis model to be trained is trained with the purpose of making the correlation coefficient output by the relationship analysis model to be trained for the training samples approach the correlation coefficient of the pre-set multiple pieces of knowledge.

Images