Intelligent guide method, device and equipment based on ERP knowledge graph and medium

Abstract

The embodiment of the invention discloses an intelligent guiding method and device based on an ERP knowledge graph, equipment and a medium, relates to the technical field of knowledge graphs, and is used for solving the problem that an existing guiding mode is highly dependent on manpower. The method comprises the following steps: preprocessing multi-source data of an ERP system to obtain standardized multi-source data; based on the data type of the standardized multi-source data, determining a corresponding knowledge extraction strategy, and extracting knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy; organizing the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and adding the knowledge graph into a graph database; processing the current user problem through a preset deep learning network to obtain entity query data; and querying the knowledge graph in the graph database based on the entity query data to obtain a display answer.

Description

Technical Field

[0001] This specification relates to the technical field of knowledge graph construction, and in particular to an intelligent wizard method, device, equipment and medium based on ERP knowledge graph. Background Art

[0002] With the continuous advancement of science and technology and the rapid development of information integration systems, data in various fields has also shown a leap-forward growth trend. The rapid development of informatization has brought many problems to enterprise data. For example, with the continuous expansion and upgrading of enterprise business systems, data is often stored in multiple different systems, databases or cloud platforms, forming data islands. This decentralized data storage method increases the difficulty of data integration and affects the accuracy and timeliness of data search. In addition, in the context of rapid data growth, the source, processing process and flow path of data have become complicated and difficult to track. Lacking an effective data traceability mechanism, enterprises often find it difficult to quickly locate the source of problems when facing data quality issues or compliance reviews, which in turn affects the timely resolution of problems.

[0003] The current system wizard mode generally relies on detailed documentation to guide user operations. Although this method lowers the threshold for using the system to a certain extent, users need to spend a lot of time reading documents and understanding system logic before they can gradually master how to use the system. Moreover, for complex systems, documents are often lengthy and difficult to understand, which increases the learning burden for users. In addition, in a rapidly changing market environment, system functions and business requirements may be adjusted frequently, while document updates often lag behind system changes, making it difficult to meet users' real-time needs. Summary of the invention

[0004] In order to solve the above technical problems, one or more embodiments of this specification provide an intelligent wizard method, device and medium based on ERP knowledge graph.

[0005] One or more embodiments of this specification adopt the following technical solutions:

[0006] One or more embodiments of this specification provide an intelligent wizard method based on an ERP knowledge graph, the method comprising:

[0007] Acquire multi-source data of the ERP system, and pre-process the multi-source data to obtain standardized multi-source data; wherein the multi-source data corresponds to each component module of the ERP system;

[0008] Determining a corresponding knowledge extraction strategy based on the data type of the standardized multi-source data, so as to extract knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy;

[0009] Organizing the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and adding the knowledge graph to a graph database;

[0010] Process the current user question through the preset deep learning network to obtain entity query data;

[0011] Based on the entity query data, the knowledge graph in the graph database is queried to obtain a display answer.

[0012] Optionally, in one or more embodiments of the present specification, the acquiring of multi-source data of the ERP system and preprocessing of the multi-source data to obtain standardized multi-source data specifically includes:

[0013] Based on the interface of each component module of the ERP system, the functional data of each component module is obtained, and multi-source data is obtained by aggregation; wherein the multi-source data includes: structured data, semi-structured data and structured data;

[0014] Performing data cleaning on the multi-source data to obtain cleaned multi-source data;

[0015] Detecting the cleaned multi-source data based on a preset hash function to remove duplicate data in the cleaned multi-source data to obtain initial multi-source data;

[0016] The initial multi-source data is matched according to a preset dictionary to implement word segmentation processing on the initial multi-source data, and the text length of the initial multi-source data after word segmentation is standardized to obtain standardized multi-source data.

[0017] Optionally, in one or more embodiments of the present specification, based on the data type of the standardized multi-source data, determining a corresponding knowledge extraction strategy to extract knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy specifically includes:

[0018] If it is determined that the standardized multi-source data is structured data, knowledge integration is performed on the standardized multi-source data to extract knowledge representation information corresponding to the standardized multi-source data according to the key-value structure of the integrated standardized multi-source data;

[0019] If it is determined that the standardized multi-source data is unstructured data or semi-structured data, the standardized multi-source data is respectively input into a preset named entity recognition model and a preset relationship extraction model;

[0020] Identify entity information of the standardized multi-source data according to the preset named entity recognition model, and identify relationship information of the standardized multi-source data according to the preset relationship extraction model;

[0021] The knowledge representation information corresponding to the standard multi-source data is determined according to the entity information and the relationship information.

[0022] Optionally, in one or more embodiments of the present specification, organizing the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and adding the knowledge graph to a graph database specifically includes:

[0023] Determine basic attributes of each node based on the entity information in the knowledge representation information, and determine basic attributes and relationships of each edge based on the relationship information;

[0024] According to the basic attributes of each node and the basic attributes and relationships of each edge, a knowledge graph corresponding to the multi-source data is constructed;

[0025] The knowledge graph is synchronized to the graph database for storage.

[0026] Optionally, in one or more embodiments of the present specification, after organizing the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and adding the knowledge graph to a graph database, the method further includes:

[0027] Determining, according to the data source of the entity information, the extended attribute corresponding to the entity information;

[0028] Querying the extended attributes and the basic attributes according to the graph database to obtain potential knowledge representation information corresponding to the knowledge graph;

[0029] The knowledge graph is updated according to the potential knowledge representation information to obtain an updated knowledge graph.

[0030] Optionally, in one or more embodiments of the present specification, the processing of the current user question by a preset deep learning network to obtain entity query data specifically includes:

[0031] Acquire sentence samples based on public channels, and annotate the sentence samples based on the preset expert interface; wherein the annotated content includes: entity information and relationship information;

[0032] Determining a replacement annotation of the annotation content according to context information of the sentence sample, so as to expand the annotation content according to the replacement annotation to obtain valid annotation content;

[0033] Inputting sentence samples with valid annotated content into a long short-term memory network, so as to train the long short-term memory network based on the sentence samples with valid annotated content to obtain a preset deep learning network;

[0034] The current user question is input into the preset deep learning network to output entity query data.

[0035] Optionally, in one or more embodiments of the present specification, querying the knowledge graph in the graph database based on the entity query data to obtain a display answer specifically includes:

[0036] Based on the entity query data and a preset SPARQL query interface, generating a query statement corresponding to the entity query data;

[0037] Query the knowledge graph in the graph database based on the query statement to obtain a query result;

[0038] The query result and the current user question are input into a preset question-answer matching model to obtain a display answer corresponding to the current user question.

[0039] One or more embodiments of this specification provide an intelligent guide device based on an ERP knowledge graph, the device comprising:

[0040] A preprocessing unit, used for acquiring multi-source data of the ERP system, and preprocessing the multi-source data to obtain standardized multi-source data; wherein the multi-source data corresponds to each component module of the ERP system;

[0041] A knowledge extraction unit, configured to determine a corresponding knowledge extraction strategy based on the data type of the standardized multi-source data, so as to extract knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy;

[0042] A graph construction unit, used for organizing the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and adding the knowledge graph to a graph database;

[0043] A processing unit, used to process the current user question through a preset deep learning network to obtain entity query data;

[0044] A query display unit is used to query the knowledge graph in the graph database based on the entity query data to obtain a display answer.

[0045] One or more embodiments of this specification provide an intelligent guide device based on an ERP knowledge graph, the device comprising:

[0046] at least one processor; and,

[0047] a memory communicatively connected to the at least one processor; wherein,

[0048] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: execute any of the above-mentioned methods.

[0049] One or more embodiments of the present specification provide a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured to execute any of the above-described methods.

[0050] At least one of the above technical solutions adopted in the embodiments of this specification can achieve the following beneficial effects:

[0051] Based on standardized multi-source data and customized knowledge extraction strategies for data types, the data-to-knowledge conversion process can be automatically completed, avoiding the traditional wizard mode that requires users to spend a lot of time to understand the system logic and data structure, which often leads to delayed processing flow. It also greatly reduces human intervention and time delays, and improves the real-time response of the system. After constructing the knowledge graph based on the extracted knowledge, it is possible to update the potential knowledge representation information based on extended attributes and basic attributes, and then continuously update the knowledge graph to adapt to the update and iteration of complex systems, ensuring the stable operation of the system. By pre-setting the deep learning network to quickly process user questions, accurate entity query data can be quickly generated, improving the response speed of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art description. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. In the drawings:

[0053] Figure 1 A schematic diagram of a flow chart of an intelligent wizard method based on ERP knowledge graph provided in an embodiment of this specification;

[0054] Figure 2 A schematic diagram of the system architecture corresponding to an intelligent wizard method based on ERP knowledge graph provided in an embodiment of this specification;

[0055] Figure 3 A schematic diagram of the structure of an intelligent guide device based on ERP knowledge graph provided in an embodiment of this specification;

[0056] Figure 4 A schematic diagram of the structure of an intelligent guide device based on an ERP knowledge graph provided in an embodiment of this specification;

[0057] Figure 5 A schematic diagram of the structure of a non-volatile storage medium provided in an embodiment of this specification. DETAILED DESCRIPTION

[0058] The embodiments of this specification provide an intelligent wizard method, device, equipment and medium based on ERP knowledge graph.

[0059] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0060] like Figure 1 As shown, the embodiment of this specification provides a flow chart of an intelligent wizard method based on an ERP knowledge graph. Figure 1 It can be seen that in one or more embodiments of this specification, an intelligent wizard method based on ERP knowledge graph includes:

[0061] S101: Acquire multi-source data of an ERP system, and pre-process the multi-source data to obtain standardized multi-source data; wherein the multi-source data corresponds to each component module of the ERP system.

[0062] In order to accurately determine the data in a complex system, the multi-source data of the ERP system is first obtained in the embodiments of this specification, wherein the multi-source data corresponds to the various component models of the ERP system, that is, structured and unstructured data are extracted from various modules of the ERP system such as finance, procurement, production, etc., and then the obtained multi-source data is pre-processed to obtain standardized multi-source data. Among them, the standard loop processing process can be to clean, remove duplicates, standardize, etc. of the extracted data to ensure data quality and consistency. This process can eliminate noise, errors, redundancy and inconsistency in the data by cleaning, removing duplicates and standardizing the multi-source data of each module in the ERP system, ensuring that subsequent analysis or decision-making is based on accurate and consistent data.

[0063] Specifically, in one or more embodiments of the present specification, multi-source data of the ERP system is obtained, and the multi-source data is pre-processed to obtain standardized multi-source data, which specifically includes the following processes:

[0064] First, based on the interface of each component module of the ERP system, the functional data of each component module is obtained, and multi-source data is obtained by aggregation; wherein, multi-source data includes: structured data, semi-structured data and structured data. Then, the multi-source data is cleaned to obtain cleaned multi-source data. Then, the cleaned multi-source data is tested according to the preset hash function to remove the duplicate data in the cleaned multi-source data and obtain the initial multi-source data. At the same time, Figure 2 The service layer of the embodiment of the present specification will match the initial multi-source data according to the preset dictionary, implement word segmentation processing on the initial multi-source data, and standardize the text length of the initial multi-source data after word segmentation to obtain standardized multi-source data.

[0065] S102: Determine a corresponding knowledge extraction strategy based on the data type of the standardized multi-source data, so as to extract knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy.

[0066] After the standardized multi-source data of each module of the ERP system is obtained based on step S102, in order to solve the problem that the traditional wizard mode takes a lot of time to understand the system logic, resulting in a large lag, in the embodiment of this specification, the corresponding knowledge extraction strategy is determined according to the data type of the standardized multi-source data, so as to extract the knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy.

[0067] In this way, different types of data have different semantic and structural characteristics. The use of targeted knowledge extraction strategies can better retain these characteristics and accurately convert them into knowledge representation information. Therefore, knowledge extraction strategies customized according to data types can more efficiently extract useful knowledge representation information from standardized multi-source data. Compared with the traditional one-size-fits-all processing method, this process can more accurately identify and process different types of data, thereby improving the efficiency and accuracy of knowledge extraction. In addition, the traditional wizard mode requires users to spend a lot of time to understand the system logic and data structure, which often leads to a lag in the processing flow. Based on standardized multi-source data and customized knowledge extraction strategies for data types, the data-to-knowledge conversion process can be automatically completed, greatly reducing human intervention and time delays, and improving the real-time response of the system.

[0068] Specifically, in one or more embodiments of the present specification, based on the data type of the standardized multi-source data, a corresponding knowledge extraction strategy is determined to extract the knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy, specifically including:

[0069] If the standardized multi-source data is determined to be structured data, then the standardized multi-source data is integrated, so as to extract the knowledge representation information corresponding to the standardized multi-source data according to the key-value structure of the integrated standardized multi-source data. If the standardized multi-source data is determined to be unstructured data or semi-structured data, then the standardized multi-source data is input into the preset named entity recognition model and the preset relationship extraction model respectively. The entity information of the standardized multi-source data is identified according to the preset named entity recognition model, and the relationship information of the standardized multi-source data is identified according to the preset relationship extraction model. The knowledge representation information corresponding to the standardized multi-source data is determined according to the entity information and the relationship information.

[0070] In the above process, for structured data, knowledge is extracted by integrating key-value structures. This method is direct and efficient, and can ensure that the required information is accurately extracted from structured data. For unstructured and semi-structured data, named entity recognition and relationship extraction models can be used to accurately identify entities in the data and the relationships between them, thereby constructing accurate knowledge representation. The entire process is highly automated and relies on preset models and algorithms, reducing the need for human intervention. This not only improves processing efficiency, but also reduces the possibility of human error. By combining entity information and relationship information, the process can generate a richer and more comprehensive knowledge representation. This knowledge representation not only contains the basic information in the data, but also reveals the inherent connection between data elements, providing strong support for subsequent data analysis, decision support, etc.

[0071] S103: Organize the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and add the knowledge graph to a graph database.

[0072] In order to build an extensible knowledge graph structure, it is convenient to quickly determine the answers to subsequent questions based on the knowledge graph structure, and solve the query difficulties caused by information islands. In the embodiment of this specification, the knowledge representation information obtained in the above step S102 will be organized to obtain the knowledge graph corresponding to the multi-source data, and the knowledge graph will be added to the graph database for storage. Specifically, in one or more embodiments of this specification, the knowledge representation information is organized to obtain the knowledge graph corresponding to the multi-source data, and the knowledge graph is added to the graph database, which specifically includes the following process: determining the basic attributes of each node based on the entity information in the knowledge representation information, and determining the basic attributes and relationships of each edge based on the relationship information. Then, according to the basic attributes of each node and the basic attributes and relationships of each edge, a knowledge graph corresponding to the multi-source data is constructed. The knowledge graph is synchronized to the graph database for storage. The nodes and edges of the knowledge graph constructed based on the above method can have rich attributes, which can be customized and expanded according to actual needs. This flexible modeling method enables the knowledge graph to adapt to various complex data scenarios and changing needs. The knowledge graph is then synchronized to the graph database, making it easier for the graph database to provide mechanisms such as cascading security permissions, thereby ensuring the security of the knowledge graph data.

[0073] Furthermore, in order to achieve complex reasoning, the knowledge graph can be continuously evolved and enriched to provide more information. In one or more embodiments of this specification, after organizing the knowledge representation information to obtain the knowledge graph corresponding to the multi-source data and adding the knowledge graph to the graph database, the method also includes the following process:

[0074] First, according to the data source of the entity information, the extended attributes corresponding to the entity information are determined. Then, the extended attributes and basic attributes are queried according to the graph database to obtain the potential knowledge representation information corresponding to the knowledge graph. Then the knowledge graph is updated according to the potential knowledge representation information to obtain the updated knowledge graph. By identifying the data source of the entity information and determining its corresponding extended attributes, the content of the knowledge graph can be further enriched. These extended attributes may include historical information, contextual information, related events, etc. of the entity, thereby providing a more comprehensive knowledge representation. The addition of extended attributes helps to verify and correct errors or omissions in the knowledge graph and improve the accuracy of the data. For example, by comparing information from different data sources, inconsistent entity attributes or relationships can be discovered and corrected. In addition. By querying the association between extended attributes and basic attributes, the potential knowledge representation information hidden in the knowledge graph can be revealed. This information may appear as new entity relationships, patterns or trends, providing strong support for data mining and insights. In complex systems, as data sources are updated and changed, knowledge graphs also need to be continuously updated and maintained. Therefore, by querying extended attributes and basic attributes, the knowledge graph can be continuously updated to adapt to the update and iteration of complex systems and ensure the stable operation of the system.

[0075] S104: Process the current user question through a preset deep learning network to obtain entity query data.

[0076] In the ERP system, data records and function modules are usually very large and complex, and there are a large number of entities and relationships in the data, and it becomes difficult to find and locate information in the system. This makes it difficult for system users to find the required information or functions when they encounter problems such as information query, business process, software function or data processing. Therefore, after obtaining the knowledge graph based on the above steps, in order to be able to use the wizard ability of the knowledge graph to help users accurately find the required information, the accurate answer to complex problems in the business scenario is realized. In the embodiment of this specification, the current user problem will be processed by the preset deep learning network to obtain entity query data. In this process, when understanding the user problem, since the acquisition of effective and rich training data requires a high cost, a semi-supervised feature learning algorithm based on context is adopted to reduce the demand for manual data annotation. Then, on the basis of part of the annotated data, the pattern and relationship of the annotated content in the sentence are learned through the long short-term memory network, and then based on the context semantics, the original annotated content is replaced with synonyms, similar words, and context semantic words, so as to obtain more and richer training data. After training the preset deep learning network according to the training data, accurate processing of the current user problem can be achieved.

[0077] Specifically, in one or more embodiments of the present specification, the current user question is processed by a preset deep learning network to obtain entity query data, which specifically includes the following process:

[0078] First, sentence samples are obtained based on public channels, and the annotation content of each sentence sample is obtained based on the preset expert interface. It should be noted that the annotation content includes: entity information and relationship information. The efficiency and accuracy of annotation are significantly improved by annotating the sentence samples through the preset expert interface. Then, the replacement annotation of the annotation content is determined according to the context information of the sentence sample, so as to expand the annotation content according to the replacement annotation to obtain the effective annotation content. The sentence samples with effective annotation content are input into the long short-term memory network, and the long short-term memory network is trained based on the sentence samples with effective annotation content to obtain the preset deep learning network. Then the current user question is input into the preset deep learning network to output the entity query data. In the annotation process, not only the entity information and relationship information in the sentence are considered, but also the replacement annotation is determined based on the context information of the sentence sample. This context-aware annotation method helps to understand the meaning and intention of the sentence more accurately. By expanding the annotation content and training the deep learning network with the expanded annotation, the network can learn more context information and language patterns, thereby improving the accuracy of the entity query data. As the amount of data increases and the annotation content expands, the performance of the preset deep learning network can be continuously improved. Then, the preset deep learning network can quickly process user questions and quickly generate accurate entity query data, thereby improving the response speed of the system.

[0079] S105: Query the knowledge graph in the graph database based on the entity query data to obtain display answers.

[0080] In the embodiment of this specification, the entity query data obtained in step S104 is used to query the knowledge graph in the graph database to obtain a display answer. Specifically, in one or more embodiments of this specification, querying the knowledge graph in the graph database based on the entity query data to obtain a display answer specifically includes the following process:

[0081] First, based on the entity query data and the preset SPARQL query interface, a query statement corresponding to the entity query data is generated. Then, the knowledge graph in the graph database is queried according to the query statement to obtain the query result. The query result and the current user question are then input into the preset question-answer matching model to obtain the display answer corresponding to the current user question. That is, SPARQL or graph query language is used to query the knowledge graph and retrieve relevant information.

[0082] like Figure 3As shown, the embodiment of this specification provides a schematic diagram of the structure of an intelligent guide device based on an ERP knowledge graph. Figure 3 It can be seen that in one or more embodiments of this specification, an intelligent guide device based on ERP knowledge graph includes:

[0083] A preprocessing unit 301 is used to obtain multi-source data of the ERP system and preprocess the multi-source data to obtain standardized multi-source data; wherein the multi-source data corresponds to each component module of the ERP system;

[0084] A knowledge extraction unit 302 is used to determine a corresponding knowledge extraction strategy based on the data type of the standardized multi-source data, so as to extract the knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy;

[0085] A graph construction unit 303 is used to organize the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and add the knowledge graph to a graph database;

[0086] The processing unit 304 is used to process the current user question through a preset deep learning network to obtain entity query data;

[0087] The query display unit 305 is used to query the knowledge graph in the graph database based on the entity query data to obtain display answers.

[0088] like Figure 4 As shown, the embodiment of this specification provides a schematic diagram of the structure of an intelligent guide device based on an ERP knowledge graph. Figure 4 It can be seen that in one or more embodiments of this specification, an intelligent guide device based on ERP knowledge graph includes:

[0089] at least one processor; and,

[0090] a memory communicatively connected to the at least one processor; wherein,

[0091] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: execute any of the above-mentioned methods.

[0092] like Figure 5 As shown, the present specification provides a schematic diagram of a non-volatile storage medium structure. Figure 5 It can be seen that in one or more embodiments of the present specification, a non-volatile storage medium stores computer-executable instructions, and the computer-executable instructions can: execute any of the methods described above.

[0093] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0094] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0095] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.

Claims

1. An intelligent wizard method based on ERP knowledge graph, It is characterized in that The method comprises: Acquire multi-source data of the ERP system, and pre-process the multi-source data to obtain standardized multi-source data; wherein the multi-source data corresponds to each component module of the ERP system; Determining a corresponding knowledge extraction strategy based on the data type of the standardized multi-source data, so as to extract the knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy; Organizing the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and adding the knowledge graph to a graph database; Process the current user question through the preset deep learning network to obtain entity query data; Based on the entity query data, the knowledge graph in the graph database is queried to obtain a display answer.

2. According to the intelligent wizard method based on ERP knowledge graph according to claim 1, It is characterized in that The step of acquiring multi-source data of the ERP system and preprocessing the multi-source data to obtain standardized multi-source data specifically includes: Based on the interface of each component module of the ERP system, the functional data of each component module is obtained, and multi-source data is obtained by aggregation; wherein the multi-source data includes: structured data, semi-structured data and structured data; Performing data cleaning on the multi-source data to obtain cleaned multi-source data; Detecting the cleaned multi-source data based on a preset hash function to remove duplicate data in the cleaned multi-source data to obtain initial multi-source data; The initial multi-source data is matched according to a preset dictionary to implement word segmentation processing on the initial multi-source data, and the text length of the initial multi-source data after word segmentation is standardized to obtain standardized multi-source data.

3. According to the intelligent guidance method based on ERP knowledge graph according to claim 1, It is characterized in that Based on the data type of the standardized multi-source data, determining a corresponding knowledge extraction strategy to extract knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy, specifically includes: If it is determined that the standardized multi-source data is structured data, knowledge integration is performed on the standardized multi-source data to extract knowledge representation information corresponding to the standardized multi-source data according to the key-value structure of the integrated standardized multi-source data; If it is determined that the standardized multi-source data is unstructured data or semi-structured data, the standardized multi-source data is respectively input into a preset named entity recognition model and a preset relationship extraction model; Identify entity information of the standardized multi-source data according to the preset named entity recognition model, and identify relationship information of the standardized multi-source data according to the preset relationship extraction model; The knowledge representation information corresponding to the standard multi-source data is determined according to the entity information and the relationship information.

4. According to the intelligent guidance method based on ERP knowledge graph according to claim 3, It is characterized in that Organizing the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and adding the knowledge graph to a graph database, specifically includes: Determine basic attributes of each node based on the entity information in the knowledge representation information, and determine basic attributes and relationships of each edge based on the relationship information; According to the basic attributes of each node and the basic attributes and relationships of each edge, a knowledge graph corresponding to the multi-source data is constructed; The knowledge graph is synchronized to the graph database for storage.

5. According to claim 4, an intelligent wizard method based on ERP knowledge graph, It is characterized in that After organizing the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and adding the knowledge graph to a graph database, the method further includes: Determining, according to the data source of the entity information, the extended attribute corresponding to the entity information; Querying the extended attributes and the basic attributes according to the graph database to obtain potential knowledge representation information corresponding to the knowledge graph; The knowledge graph is updated according to the potential knowledge representation information to obtain an updated knowledge graph.

6. According to claim 1, an intelligent wizard method based on ERP knowledge graph, It is characterized in that The process of processing the current user question through the preset deep learning network to obtain entity query data specifically includes: Acquire sentence samples based on public channels, and annotate the sentence samples based on the preset expert interface; wherein the annotated content includes: entity information and relationship information; Determining a replacement annotation of the annotation content according to context information of the sentence sample, so as to expand the annotation content according to the replacement annotation to obtain valid annotation content; Inputting sentence samples with valid annotated content into a long short-term memory network, so as to train the long short-term memory network based on the sentence samples with valid annotated content to obtain a preset deep learning network; The current user question is input into the preset deep learning network to output entity query data.

7. According to claim 1, an intelligent wizard method based on ERP knowledge graph, It is characterized in that The knowledge graph in the graph database is queried based on the entity query data to obtain a display answer, specifically including: Based on the entity query data and a preset SPARQL query interface, generating a query statement corresponding to the entity query data; Query the knowledge graph in the graph database based on the query statement to obtain a query result; The query result and the current user question are input into a preset question-answer matching model to obtain a display answer corresponding to the current user question.

8. An intelligent guide device based on ERP knowledge graph, It is characterized in that The device comprises: A preprocessing unit, used for acquiring multi-source data of the ERP system, and preprocessing the multi-source data to obtain standardized multi-source data; wherein the multi-source data corresponds to each component module of the ERP system; A knowledge extraction unit, configured to determine a corresponding knowledge extraction strategy based on the data type of the standardized multi-source data, so as to extract knowledge representation information corresponding to the standardized multi-source data according to the knowledge extraction strategy; A graph construction unit, used for organizing the knowledge representation information to obtain a knowledge graph corresponding to the multi-source data, and adding the knowledge graph to a graph database; A processing unit, used to process the current user question through a preset deep learning network to obtain entity query data; A query display unit is used to query the knowledge graph in the graph database based on the entity query data to obtain a display answer.

9. An intelligent guide device based on ERP knowledge graph, It is characterized in that The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: execute any of the methods described in claims 1-7.

10. A non-volatile storage medium storing computer executable instructions, It is characterized in that The computer executable instructions can: execute the method described in any one of claims 1 to 7.

Images