Intelligent construction method and system for IT operation and maintenance knowledge base fusing knowledge graph

Abstract

The invention discloses a knowledge graph fused IT operation and maintenance knowledge base intelligent construction method and system. The method comprises the steps of preprocessing collected original data; an initial entity set and a relation set of the knowledge graph are constructed based on the standardized operation and maintenance data set, entities comprise equipment, components, fault types and solutions, and relations comprise association between the equipment and the components and association between the fault types and the solutions; performing semantic analysis on text information in the operation and maintenance data set, extracting key information as attribute information of the knowledge graph, and associating the key information with corresponding entities or relationships; optimizing the knowledge graph according to a preset algorithm, including entity alignment, relation reasoning and knowledge fusion, so as to generate a complete IT operation and maintenance knowledge graph; and storing the constructed IT operation and maintenance knowledge graph in a knowledge base. According to the method, the quality and intelligence of the knowledge graph are effectively improved, the problems of entity redundancy, insufficient excavation of implicit relationships and knowledge repetition are solved, the construction efficiency of the IT operation and maintenance knowledge base and the quality of a basic structure are remarkably improved, and a solid foundation is laid for subsequent knowledge optimization, reasoning and application.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of knowledge base operation and maintenance, and in particular to an IT operation and maintenance knowledge base intelligent construction method and system integrating knowledge graphs. BACKGROUND

[0002] In the field of modern IT operation and maintenance, with the rapid development of information technology and the increasing complexity of systems, traditional operation and maintenance methods face many challenges. Traditional operation and maintenance knowledge management mainly relies on manual recording and experience summary, which has the following problems:

[0003] Operation and maintenance data is scattered in multiple heterogeneous data sources, such as system logs, device operation data, fault work order records, etc., making it difficult to form a unified knowledge system. Knowledge update is not timely: new knowledge generated during operation and maintenance cannot be integrated into the knowledge base in a timely manner, resulting in insufficient timeliness and accuracy of the knowledge base. Lack of intelligent analysis capability: traditional knowledge bases cannot perform deep mining and correlation analysis on operation and maintenance data, making it difficult to discover potential fault patterns and solutions.

[0004] In order to solve these problems, knowledge graph technology has been gradually introduced into the field of IT operation and maintenance. Knowledge graph is a structured semantic knowledge base that can represent entities and their relationships in the form of a graph, providing strong support for knowledge storage, management and query. However, existing knowledge graph construction methods still have some shortcomings in the application of IT operation and maintenance, such as: the optimization and updating mechanism of the knowledge graph is not perfect, and it is difficult to adapt to the rapidly changing knowledge needs in the IT operation and maintenance environment. The construction method of the knowledge graph lacks deep semantic understanding and correlation analysis of operation and maintenance data, resulting in insufficient accuracy and completeness of the knowledge. Therefore, a more efficient, intelligent and deep mining of operation and maintenance data IT operation and maintenance knowledge base construction method is needed to meet the needs of modern IT operation and maintenance. SUMMARY

[0005] The present application provides an IT operation and maintenance knowledge base intelligent construction method and system integrating knowledge graphs, which solves the above technical problems pointed out in the prior art.

[0006] The present application provides an IT operation and maintenance knowledge base intelligent construction method integrating knowledge graphs, comprising the following operation steps:

[0007] Collecting IT operation and maintenance related raw data from multiple heterogeneous data sources, including system logs, device operation data, fault work order records and user feedback information;

[0008] Preprocessing the collected raw data, including data cleaning, format unification and data labeling, to form a standardized operation and maintenance data set;

[0009] An initial entity set and a relationship set of the knowledge graph are constructed based on a standardized operation and maintenance dataset, wherein the entities include devices, components, fault types, and solutions, and the relationships include the association between devices and components and the association between fault types and solutions;

[0010] Text information in the operation and maintenance dataset is subjected to semantic analysis, key information is extracted as attribute information of the knowledge graph, and the key information is associated with corresponding entities or relationships;

[0011] The knowledge graph is optimized according to a preset algorithm, including entity alignment, relationship reasoning, and knowledge fusion, to generate a complete IT operation and maintenance knowledge graph;

[0012] The constructed IT operation and maintenance knowledge graph is stored in a knowledge base.

[0013] Preferably, the specific steps of the data cleaning include:

[0014] The collected original data is subjected to format recognition to determine whether the data conforms to a preset data format specification;

[0015] For data that does not conform to the format specification, a format conversion operation is performed to convert the data into a unified format;

[0016] The converted data is subjected to missing value detection, and for data having missing values, a preset filling strategy is used for filling;

[0017] The abnormal values in the data are identified and processed, including deleting the abnormal values or replacing the abnormal values with reasonable values;

[0018] The data is subjected to a deduplication operation to remove duplicate data records.

[0019] Preferably, the preset filling strategy includes:

[0020] For missing values of numerical data, mean filling, median filling, or mode filling is used for filling;

[0021] For missing values of categorical data, mode filling or similarity-based filling is used for filling;

[0022] For missing values of text data, a preset default text or context-based generated text is used for filling.

[0023] Preferably, the specific steps of constructing the initial entity set and the relationship set of the knowledge graph include:

[0024] The standardized operation and maintenance dataset is subjected to entity recognition to extract entity information of devices, components, fault types, and solutions;

[0025] Classify and label the extracted entity information according to the preset entity type and attribute definitions;

[0026] Analyze the relationships between entities in the operation and maintenance data set, including the composition relationship between equipment and components, the relationship between fault types and equipment, and the correspondence between solutions and fault types;

[0027] Based on the identified entities and relationships, the initial entity set and relationship set of the knowledge graph are constructed.

[0028] Preferably, the specific steps of analyzing the association relationship between entities include:

[0029] For the association between devices and components, the components included in the device are determined by parsing the device configuration information and component installation records;

[0030] For the correlation between fault types and equipment, analyze the causal relationship between the fault and the equipment based on the equipment operation data and fault logs at the time of the fault;

[0031] As for the correspondence between solutions and faults, the matching relationship between solutions and faults is determined by analyzing the fault work order records and the implementation effects of the solutions.

[0032] Preferably, the specific steps of optimizing the knowledge graph include:

[0033] Align entities in the knowledge graph, calculate the similarity between entities, and merge or associate entities with higher similarity;

[0034] Use the preset relational reasoning algorithm to reason about the relationships in the knowledge graph, discover potential implicit relationships, and add them to the knowledge graph;

[0035] Fuse the knowledge in the knowledge graph, analyze the correlation between knowledge from different sources, and integrate and optimize duplicate or redundant knowledge.

[0036] Preferably, the formula for calculating entity similarity is:

[0037] ;

[0038] in, and are the two entities to be compared, is the number of attributes of the entity, For the The similarity calculation result of each attribute is calculated as follows:

[0039] ;

[0040] wherein, and are the word frequencies on the first and the second attributes, respectively.

[0041] The application provides an intelligent construction system of an IT operation and maintenance knowledge base fused with a knowledge graph, comprising a collection module, a preprocessing module, a construction module, an association module, an optimization module, and a storage module.

[0042] The collection module is configured to collect original data related to IT operation and maintenance from a plurality of heterogeneous data sources, wherein the heterogeneous data sources include system logs, device operation data, fault work order records, and user feedback information.

[0043] The preprocessing module is configured to preprocess the collected original data, including data cleaning, format unification, and data labeling, to form a standardized operation and maintenance data set.

[0044] The construction module is configured to construct an initial entity set and a relationship set of the knowledge graph based on the standardized operation and maintenance data set, wherein the entities include devices, components, fault types, and solutions, and the relationships include the association between devices and components and the association between fault types and solutions.

[0045] The association module is configured to perform semantic analysis on the text information in the operation and maintenance data set, extract key information as attribute information of the knowledge graph, and associate the attribute information with corresponding entities or relationships.

[0046] The optimization module is configured to optimize the knowledge graph according to a preset algorithm, including entity alignment, relationship reasoning, and knowledge fusion, to generate a complete IT operation and maintenance knowledge graph.

[0047] The storage module is configured to store the constructed IT operation and maintenance knowledge graph into a knowledge base.

[0048] A storage medium has a computer program stored therein, and the computer program is executed by a processor to implement the steps of the above-mentioned intelligent construction method of an IT operation and maintenance knowledge base fused with a knowledge graph.

[0049] Compared with the prior art, the embodiments of the application have at least the following technical advantages:

[0050] An analysis of the method for intelligently constructing an IT operation and maintenance knowledge base integrated with a knowledge graph provided by the present invention shows that, in specific applications, first, raw data related to IT operation and maintenance is collected from multiple heterogeneous data sources, including system logs, equipment operation data, fault work order records, and user feedback information; the collected raw data is preprocessed, including data cleaning, format unification, and data labeling, to form a standardized operation and maintenance data set;

[0051] Then, based on the standardized operation and maintenance dataset, we construct the initial entity and relationship sets of the knowledge graph. Entities include equipment, components, fault types, and solutions, while relationships include the association between equipment and components, and the association between fault types and solutions. We perform semantic parsing on the text information in the operation and maintenance dataset, extract key information as attribute information for the knowledge graph, and associate it with the corresponding entities or relationships.

[0052] Optimize the knowledge graph according to preset algorithms, including entity alignment, relationship reasoning, and knowledge fusion, to generate a complete IT operation and maintenance knowledge graph; store the constructed IT operation and maintenance knowledge graph in the knowledge base;

[0053] The above technical solution collects data from multiple heterogeneous data sources (such as system logs, device operation data, and troubleshooting work ticket records), cleans it, standardizes its format, and annotates it to form a standardized operation and maintenance dataset. This process improves data quality and consistency, providing a reliable data foundation for the subsequent construction of a knowledge graph. Semantic parsing technology then extracts key information from the operation and maintenance data as knowledge graph attributes and associates them with entities or relationships. This process strengthens the semantic expressiveness of the knowledge graph. Optimization operations such as entity alignment, relationship reasoning, and knowledge fusion enhance the integrity and accuracy of the knowledge graph. Entity and relationship analysis, by parsing device configuration information, operation data, and fault logs, deeply explores causal and matching relationships between devices and components, faults and devices, and solutions and faults, further enhancing the expressive power of the knowledge graph. Relational reasoning algorithms discover potential implicit relationships, and knowledge fusion integrates redundant knowledge from different sources, enhancing the dynamic expansion and practicality of the knowledge base.

[0054] The above technical methods are implemented through multiple technologies such as data quality improvement, structured knowledge expression, semantic association, optimized reasoning and dynamic expansion, and have built a comprehensive and efficient IT operation and maintenance knowledge base. Its technical effects have significantly enhanced the knowledge management capabilities and intelligence level in the IT operation and maintenance field. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the specific embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the specific embodiments or prior art description. Obviously, the drawings described below are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0056] Figure 1 A fusion knowledge graph of the IT operation and maintenance knowledge base intelligent construction method provided by the embodiment one of the present application is shown in the following steps.

[0057] Figure 2 A fusion knowledge graph of the IT operation and maintenance knowledge base intelligent construction system provided by the embodiment two of the present application is shown in the following steps.

[0058] Figure 3 A storage medium provided by the embodiment three of the present application is shown in the following steps.

[0059] Mark: acquisition module 10; preprocessing module 20; construction module 30; correlation module 40; optimization module 50; storage module 60; processor 1110; communication interface 1120; memory 1130; computer storage medium 1140. DETAILED DESCRIPTION

[0060] The technical solutions of the present application will be described clearly and completely below in combination with the drawings. Obviously, the described embodiments are some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor belong to the protection scope of the present application.

[0061] The present application will be described in further detail below through specific embodiments and in combination with the drawings.

[0062] Embodiment one

[0063] As shown in the following steps, the present application proposes a fusion knowledge graph of the IT operation and maintenance knowledge base intelligent construction method, including the following operation steps: Figure 1

[0064] Step S101: collecting IT operation and maintenance related raw data from multiple heterogeneous data sources, the heterogeneous data sources including system logs, device operation data, fault work order records and user feedback information;

[0065] Step S102: preprocessing the collected raw data, including data cleaning, format unification and data labeling, to form a standardized operation and maintenance data set;

[0066] ​Step S103: Based on the standardized operation and maintenance dataset, an initial entity set and a relationship set of the knowledge graph are constructed, wherein the entities include devices, components, fault types, solutions, and the relationships include the association between devices and components and the association between fault types and solutions.

[0067] Step S104: The text information in the operation and maintenance dataset is subjected to semantic analysis, key information is extracted as attribute information of the knowledge graph, and the key information is associated with corresponding entities or relationships.

[0068] Step S105: The knowledge graph is optimized according to a preset algorithm, including entity alignment, relationship reasoning, and knowledge fusion, to generate a complete IT operation and maintenance knowledge graph.

[0069] Step S106: The constructed IT operation and maintenance knowledge graph is stored in a knowledge base, and knowledge query, update, and maintenance interfaces are provided.

[0070] Data collection: Collecting IT operation and maintenance related raw data from multiple heterogeneous data sources. For example, obtaining system logs from servers through log collection tools, obtaining device operation data through device management interfaces, obtaining fault work order records from operation and maintenance management systems, and obtaining user feedback information through user feedback channels. The collected data is stored in a distributed storage system for subsequent processing.

[0071] Data preprocessing: Identifying the format of the collected raw data and determining whether the data conforms to the preset data format specification. For example, checking whether the log data conforms to a specific log format and whether the device operation data contains necessary fields.

[0072] For data that does not conform to the format specification, perform format conversion to convert it to a unified format. For example, convert different format log data to a unified JSON format.

[0073] Detect missing values in the converted data, and fill in missing values using a preset filling strategy. For example, fill in missing values of numerical data with the mean value, and fill in missing values of categorical data with the mode.

[0074] Identify and handle outliers in the data, including deleting outliers or replacing them with reasonable values. For example, correct abnormal high or low values in device operation data by comparing with historical data.

[0075] Perform data deduplication to remove duplicate data records. For example, detect and delete duplicate log records using a hash algorithm.

[0076] Data Annotation: After cleaning the data, annotate the key information in the data, such as device name, fault type, solution, etc. For example, use regular expressions or machine learning models to annotate fault information in logs.

[0077] Entity Recognition and Classification: Perform entity recognition on the standardized operation and maintenance dataset, extracting entity information such as devices, components, fault types, and solutions. For example, use named entity recognition algorithms in natural language processing techniques to extract device names and fault types from text data.

[0078] According to the predefined entity type and attribute definition, classify and annotate the extracted entity information. For example, classify device entities into server, network device, storage device, etc., and annotate their attributes such as device model, IP address, etc.

[0079] Relationship Identification and Construction: Analyze the association between entities in the operation and maintenance dataset, including the composition relationship between devices and components, the association between faults and devices, and the corresponding relationship between solutions and faults. For example, by analyzing the configuration information of devices and the installation records of components, determine the components contained in the device; by analyzing the device running data and fault logs when the fault occurs, determine the causal relationship between the fault and the device.

[0080] Based on the identified entities and associated relationships, construct the initial entity set and relationship set of the knowledge graph. For example, connect device entities and component entities through the "composition" relationship, and connect fault entities and solution entities through the "corresponding" relationship.

[0081] Attribute Information Extraction and Association: Use natural language processing techniques to perform semantic analysis on the text information in the operation and maintenance dataset, extracting key information as attribute information of the knowledge graph. For example, extract fault description, solution description, etc. from fault work order records, and through semantic analysis, extract key information such as fault cause, solution steps, etc.

[0082] Associate the extracted attribute information with the corresponding entity or relationship. For example, associate the fault cause attribute with the fault entity, and associate the solution step attribute with the solution entity.

[0083] Knowledge Graph Optimization:

[0084] Entity Alignment: Perform alignment operations on entities in the knowledge graph, and by calculating the similarity between entities, merge or associate entities with high similarity. For example, by calculating the similarity of entity names, attribute values, etc., merge entities representing the same device in different data sources.

[0085] The formula for calculating entity similarity is:

[0086] ;

[0087] wherein, and are two entities to be compared, is the number of attributes of the entities, is the similarity calculation result of the th attribute, and the specific calculation formula is:

[0088] ;

[0089] wherein, and are the word frequencies of the entities and on the th attribute.

[0090] The preset relationship reasoning algorithm is used to reason the relationships in the knowledge graph, discover potential implicit relationships, and add them to the knowledge graph. For example, through the path reasoning algorithm, if entity A has a relationship with entity B, and entity B has a relationship with entity C, it is inferred that there may be a certain relationship between entity A and entity C.

[0091] Knowledge fusion: fuse the knowledge in the knowledge graph, and integrate and optimize the repeated or redundant knowledge by analyzing the relevance between different sources of knowledge. For example, different solutions to the same fault in different data sources are combined to extract the optimal solution.

[0092] The above technical solution collects data from multiple heterogeneous data sources (such as system logs, device running data, fault work order records, etc.), and cleans, formats, and data labels them to form a standardized operation and maintenance data set. This process improves the quality and consistency of the data, providing a reliable data foundation for the subsequent construction of the knowledge graph. Then through semantic analysis technology, the key information in the operation and maintenance data is extracted as the attribute information of the knowledge graph, and is associated with entities or relationships. This process enhances the semantic expression ability of the knowledge graph. Through entity alignment, relationship reasoning and knowledge fusion optimization operations, the completeness and accuracy of the knowledge graph are enhanced. In entity and relationship analysis, by analyzing device configuration information, running data, fault logs, etc., the causal and matching relationships between devices and components, faults and devices, solutions and faults are further excavated, further enhancing the expression ability of the knowledge graph. Through the relationship reasoning algorithm, potential implicit relationships are discovered, and redundant knowledge from different sources is integrated through knowledge fusion, enhancing the dynamic expansion ability and practicality of the knowledge base.

[0093] The technical method has the technical effects of improving data quality, structuring knowledge expression, semantic association, optimizing reasoning, and dynamic expansion, and builds a comprehensive and efficient IT operation knowledge base, which significantly enhances the knowledge management capability and intelligence level in the field of IT operation.

[0094] Preferably, the specific steps of data cleaning in step S102 include:

[0095] Step S201: Format recognition is performed on the collected raw data to determine whether the data conforms to the preset data format specification;

[0096] Step S202: For data that does not conform to the format specification, format conversion is performed to convert it to a unified format;

[0097] Step S203: Missing value detection is performed on the converted data, and for data with missing values, a preset filling strategy is used for filling;

[0098] Step S204: Abnormal values in the data are identified and processed, including deleting abnormal values or replacing them with reasonable values;

[0099] Step S205: Data is subjected to a de-duplication operation to remove duplicate data records.

[0100] The converted data is subjected to missing value detection, and for data with missing values, a preset filling strategy is used for filling; abnormal values in the data are identified and processed, including deleting abnormal values or replacing them with reasonable values. For example, for abnormal high or low values in device operation data, correction is performed by comparing with historical data. Data is subjected to a de-duplication operation to remove duplicate data records. For example, duplicate log records are detected and deleted by using a hash algorithm. The above operations are all specific steps of data cleaning, and will not be described again.

[0101] Preferably, the preset filling strategy in step S203 includes:

[0102] Step S301: For missing values of numerical data, mean filling, median filling, or mode filling is used for filling;

[0103] Step S302: For missing values of categorical data, mode filling or similarity-based filling is used for filling;

[0104] Step S303: For missing values of text data, a preset default text or context-based generated text is used for filling.

[0105] In specific technical solutions, the filling strategies of different types of data (such as mean value filling of numerical data and context generation of text data) adapt to diversified data requirements. The above execution steps process missing values through various preset filling strategies (steps S203 and S301-S303), ensuring the integrity and consistency of the data.

[0106] Preferably, the specific steps of constructing the initial entity set and the relationship set of the knowledge graph in step S103 include:

[0107] Step S401: performing entity recognition on the standardized operation and maintenance data set to extract entity information of devices, components, fault types, and solutions;

[0108] Step S402: classifying and labeling the extracted entity information according to the preset entity type and attribute definition;

[0109] Step S403: analyzing the association relationships between entities in the operation and maintenance data set, including the composition relationship between devices and components, the association relationship between fault types and devices, and the corresponding relationship between solutions and fault types;

[0110] Step S404: constructing the initial entity set and the relationship set of the knowledge graph based on the recognized entities and association relationships.

[0111] In specific technical solutions, through entity recognition of the standardized operation and maintenance data set (step S401), key entities such as devices, components, fault types, and solutions are systematically extracted, laying a foundation for the construction of the knowledge graph. According to the preset entity type and attribute definition (step S402), the extracted entity information is classified and labeled, making the definition of entities more standardized. By analyzing the composition relationship between devices and components, the association relationship between fault types and devices, and the corresponding relationship between solutions and fault types (step S403), important associations between entities in the operation and maintenance data are mined and clarified.

[0112] Therefore, the relationship network of the knowledge graph is constructed, enabling the knowledge graph to effectively express the complex associations between entities, thus being more consistent with actual operation and maintenance scenarios.

[0113] Finally, based on the recognized entities and association relationships (step S404), the construction of the initial entity set and the relationship set of the knowledge graph is completed, forming a preliminary knowledge structure containing devices, components, fault types, solutions, and their associations.

[0114] Preferably, the specific steps of analyzing the association relationships between entities in step S403 include:

[0115] Step S501: For the association relationship between the device and the component, the components contained in the device are determined by analyzing the configuration information of the device and the installation record of the component;

[0116] Step S502: For the association relationship between the fault type and the device, the causal relationship between the fault and the device is analyzed according to the device operation data and the fault log at the time of the fault;

[0117] Step S503: For the corresponding relationship between the solution and the fault, the matching relationship between the solution and the fault is determined by analyzing the fault work order record and the implementation effect of the solution.

[0118] In the specific technical solution, the causal and matching relationships between entities are established by analyzing the device operation data, the fault log and the implementation effect of the solution (steps S501-S503). The deep association analysis between entities makes the knowledge graph closer to the actual operation and maintenance scene.

[0119] Preferably, the specific steps of knowledge graph optimization in step S105 include:

[0120] Step S601: Aligning the entities in the knowledge graph, merging or associating the entities with high similarity by calculating the similarity between the entities;

[0121] Step S602: Using a preset relationship reasoning algorithm to reason the relationships in the knowledge graph, discovering potential implicit relationships and adding them to the knowledge graph;

[0122] Step S603: Fusing the knowledge in the knowledge graph, integrating and optimizing the repeated or redundant knowledge by analyzing the associativity between different sources of knowledge.

[0123] In the specific technical solution, the similarity between entities is calculated by entity alignment (step S601), and the entities with high similarity are merged or associated to avoid entity repetition or redundancy. Then, the preset relationship reasoning algorithm (step S602) is used to mine potential implicit relationships and add them to the knowledge graph. Finally, the associativity between different sources of knowledge is analyzed (step S603) to integrate and optimize the repeated or redundant knowledge.

[0124] Preferably, the formula for calculating the similarity between entities in step S601 is:

[0125] ;

[0126] wherein, and are two entities to be compared, is the number of attributes of the entity, is the first The similarity calculation result of the attributes, and the specific calculation formula is:

[0127] ;

[0128] Wherein, and The word frequency of the entity and On the first attribute.

[0129] Embodiment two

[0130] Referring to Figure 2 , the application provides an IT operation and maintenance knowledge base intelligent construction system based on a knowledge graph, comprising: a collection module 10, a preprocessing module 20, a construction module 30, an association module 40, an optimization module 50 and a storage module 60.

[0131] The collection module is used for collecting original data related to IT operation and maintenance from a plurality of heterogeneous data sources, wherein the heterogeneous data sources include system logs, device operation data, fault work order records and user feedback information.

[0132] The preprocessing module is used for preprocessing the collected original data, including data cleaning, format unification and data labeling, to form a standardized operation and maintenance data set.

[0133] The construction module is used for constructing an initial entity set and a relationship set of the knowledge graph based on the standardized operation and maintenance data set, wherein the entities include devices, components, fault types and solutions, and the relationships include the association between devices and components and the association between fault types and solutions.

[0134] The association module is used for performing semantic analysis on the text information in the operation and maintenance data set, extracting key information as attribute information of the knowledge graph, and associating the attribute information with the corresponding entities or relationships.

[0135] The optimization module is used for optimizing the knowledge graph according to a preset algorithm, including entity alignment, relationship reasoning and knowledge fusion, to generate a complete IT operation and maintenance knowledge graph.

[0136] The storage module is used for storing the constructed IT operation and maintenance knowledge graph into the knowledge base.

[0137] Embodiment three

[0138] As Figure 3 ​As shown, in another aspect, the present embodiment three is based on the intelligent construction method of the IT operation and maintenance knowledge base fused with the knowledge graph provided in the present embodiment one, and further provides a computer storage medium 1140 (referred to as a storage medium for short). A structure framework schematic diagram of the computer storage medium provided for the present embodiment three includes:

[0139] a memory 1130 for storing a computer program;

[0140] a communication interface 1120 for connecting the memory 1130 with the processor 1110;

[0141] the processor 1110 is configured to execute the computer program, and realize the intelligent construction method of the IT operation and maintenance knowledge base fused with the knowledge graph in the combination disclosed in any of the above embodiments.

[0142] It can be understood that the embodiments described herein can be realized by hardware, software, firmware, middleware, microcode or a combination thereof. For hardware implementation, the processing unit can be realized in one or more application specific integrated circuits, digital signal processors, digital signal processing devices, programmable logic devices, field programmable gate arrays, general processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions described in the present application, or a combination thereof.

[0143] For software implementation, the technologies described herein can be realized by units performing the functions described herein. The software code can be stored in the memory and executed by the processor. The memory can be implemented in the processor or outside the processor.

[0144] In summary, the present embodiment provides an intelligent construction method and system of the IT operation and maintenance knowledge base fused with the knowledge graph. The method combines entity recognition, classification annotation, relationship analysis and initial construction, and quickly and efficiently constructs an initial knowledge graph with comprehensive coverage and clear logic. By analyzing the relevance between different sources of knowledge (step S603), the repeated or redundant knowledge is integrated and optimized, ensuring that the content of the knowledge graph is more refined and efficient, and avoiding information conflicts. The optimization scheme effectively improves the quality and intelligence of the knowledge graph, and solves the problems of entity redundancy, insufficient implicit relationship mining and knowledge repetition.

[0145] The intelligent construction method of the IT operation and maintenance knowledge base fused with the knowledge graph provided by the present embodiment significantly improves the construction efficiency and the quality of the basic structure of the IT operation and maintenance knowledge base, and lays a solid foundation for subsequent knowledge optimization, reasoning and application.

[0146] Those skilled in the art can clearly understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software manner depends on the specific application and design constraints of the technical solution. The skilled person can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0147] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0148] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; the skilled in the art can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for intelligently constructing an IT operation and maintenance knowledge base integrating a knowledge graph, characterized in that: The following steps are included: Collecting raw data related to IT operations and maintenance from multiple heterogeneous data sources, including system logs, equipment operation data, fault ticket records, and user feedback information; Preprocess the collected raw data, including data cleaning, format unification, and data labeling, to form a standardized operation and maintenance data set; Build the initial entity and relationship sets of the knowledge graph based on the standardized operation and maintenance dataset. Entities include equipment, components, fault types, and solutions, while relationships include the association between equipment and components, and the association between fault types and solutions. Perform semantic analysis on the text information in the operation and maintenance dataset, extract key information as attribute information of the knowledge graph, and associate it with the corresponding entities or relationships; Optimize the knowledge graph based on preset algorithms, including entity alignment, relationship reasoning, and knowledge fusion, to generate a complete IT operations knowledge graph; Store the constructed IT operation and maintenance knowledge graph in the knowledge base.

2. The method for intelligently constructing an IT operation and maintenance knowledge base integrating a knowledge graph according to claim 1 is characterized in that: The specific steps of data cleaning include: Perform format recognition on the collected raw data to determine whether the data conforms to the preset data format specifications; For data that does not conform to the format specifications, perform format conversion operations to convert it into a unified format; Perform missing value detection on the converted data, and use the preset filling strategy to fill in the data with missing values; Identify and handle outliers in the data, including deleting them or replacing them with reasonable values; Perform deduplication operations on the data to remove duplicate data records.

3. The method for intelligently constructing an IT operation and maintenance knowledge base integrating a knowledge graph according to claim 2 is characterized in that: The preset filling strategies include: For missing values ​​of numerical data, fill them with mean, median or mode. For missing values ​​of categorical data, mode filling or similarity-based filling is used; For missing values ​​in text data, preset default text or generated text based on the context is used to fill them.

4. The method for intelligently constructing an IT operation and maintenance knowledge base integrating a knowledge graph according to claim 1 is characterized in that: The specific steps of constructing the initial entity set and relationship set of the knowledge graph include: Perform entity recognition on standardized operation and maintenance data sets to extract entity information of equipment, components, fault types, and solutions; Classify and label the extracted entity information according to the preset entity type and attribute definitions; Analyze the relationships between entities in the operation and maintenance data set, including the composition relationship between equipment and components, the relationship between fault types and equipment, and the correspondence between solutions and fault types; Based on the identified entities and relationships, the initial entity set and relationship set of the knowledge graph are constructed.

5. The method for intelligently constructing an IT operation and maintenance knowledge base integrating a knowledge graph according to claim 4 is characterized in that: The specific steps of analyzing the association relationship between entities include: For the association between devices and components, the components included in the device are determined by parsing the device configuration information and component installation records; For the correlation between fault types and equipment, analyze the causal relationship between the fault and the equipment based on the equipment operation data and fault logs at the time of the fault; As for the correspondence between solutions and faults, the matching relationship between solutions and faults is determined by analyzing the fault work order records and the implementation effects of the solutions.

6. The method for intelligently constructing an IT operation and maintenance knowledge base integrating a knowledge graph according to claim 1 is characterized in that: The specific steps of the knowledge graph optimization include: Align entities in the knowledge graph, calculate the similarity between entities, and merge or associate entities with higher similarity; Use the preset relational reasoning algorithm to reason about the relationships in the knowledge graph, discover potential implicit relationships, and add them to the knowledge graph; Fuse the knowledge in the knowledge graph, analyze the correlation between knowledge from different sources, and integrate and optimize duplicate or redundant knowledge.

7. The method for intelligently constructing an IT operation and maintenance knowledge base integrating a knowledge graph according to claim 6 is characterized in that: The formula for calculating entity similarity is: ； in, and are the two entities to be compared, is the number of attributes of the entity, For the The similarity calculation result of each attribute is calculated as follows: ； in, and Entity and In the The word frequency on an attribute.

8. An intelligent construction system for IT operation and maintenance knowledge base integrated with knowledge graph, comprising: Acquisition module, preprocessing module, construction module, association module, optimization module and storage module; The acquisition module is used to collect raw data related to IT operations and maintenance from multiple heterogeneous data sources, including system logs, equipment operation data, fault ticket records, and user feedback information; The preprocessing module is used to preprocess the collected raw data, including data cleaning, format unification, and data labeling to form a standardized operation and maintenance data set; A construction module is used to construct the initial entity set and relationship set of the knowledge graph based on the standardized operation and maintenance dataset. The entities include equipment, components, fault types, and solutions, and the relationships include the association between equipment and components, and the association between fault types and solutions. The association module is used to perform semantic analysis on the text information in the operation and maintenance dataset, extract key information as attribute information of the knowledge graph, and associate it with the corresponding entities or relationships; The optimization module is used to optimize the knowledge graph according to the preset algorithm, including entity alignment, relationship reasoning, and knowledge fusion, to generate a complete IT operation and maintenance knowledge graph; The storage module is used to store the constructed IT operation and maintenance knowledge graph in the knowledge base.

9. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for intelligently constructing an IT operation and maintenance knowledge base integrating a knowledge graph as described in any one of claims 1 to 7.

Images