CN118964682A - Intelligent construction method of software knowledge graph based on meta-level low-code platform - Google Patents
Intelligent construction method of software knowledge graph based on meta-level low-code platform Download PDFInfo
- Publication number
- CN118964682A CN118964682A CN202411449597.8A CN202411449597A CN118964682A CN 118964682 A CN118964682 A CN 118964682A CN 202411449597 A CN202411449597 A CN 202411449597A CN 118964682 A CN118964682 A CN 118964682A
- Authority
- CN
- China
- Prior art keywords
- target
- knowledge
- code
- data
- knowledge graph
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/901—Indexing; Data structures therefor; Storage structures
- G06F16/9024—Graphs; Linked lists
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, which relates to the technical field of knowledge graph construction and comprises the following steps: extracting target knowledge elements from different types of data sources, integrating the target knowledge elements into a target meta-model, and establishing a target meta-data management system; based on a preset graph database, realizing the mapping construction of the low-code development knowledge to obtain a target knowledge graph; and monitoring the change state of codes in the code warehouse in real time, and dynamically updating the target knowledge graph according to the change state data. Integrating the target knowledge elements into a meta model by extracting the target knowledge elements; based on a preset graph database, realizing the mapping construction of low-code development knowledge; the method monitors the change state of codes in the code warehouse in real time, dynamically updates the knowledge graph according to the change state data, can support rapid construction and updating of the software knowledge graph, is applied to various low-code development platforms, and improves development efficiency and software quality.
Description
Technical Field
The invention relates to the technical field of knowledge graph construction, in particular to an intelligent software knowledge graph construction method based on a metalevel low-code platform.
Background
In the context of current digital transformation, enterprises are increasingly demanding for intelligence, automation and integration. The software knowledge graph is used as a key technology for connecting different systems and data, and has wide application prospect. However, the conventional software knowledge graph construction method often has the problems of long development period, high technical threshold, high maintenance cost and the like, and is difficult to meet the requirement of enterprises for quick response to market changes. Therefore, how to quickly construct and update the knowledge graph of the software becomes one of the current research centers.
Therefore, the invention provides an intelligent software knowledge graph construction method based on a metalevel low-code platform, which can be applied to various low-code development platforms, supports quick construction and updating of the software knowledge graph, and improves development efficiency and software quality.
Disclosure of Invention
The invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, which is used for integrating a meta model by extracting target knowledge elements; based on a preset graph database, realizing the mapping construction of low-code development knowledge; the method monitors the change state of codes in the code warehouse in real time, dynamically updates the knowledge graph according to the change state data, can support rapid construction and updating of the software knowledge graph, is applied to various low-code development platforms, and improves development efficiency and software quality.
The invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, which comprises the following steps:
Step 1: extracting target knowledge elements from different types of data sources, integrating the target knowledge elements into a target meta-model, and establishing a target meta-data management system;
Step 2: based on a preset graph database, realizing the mapping construction of the low-code development knowledge to obtain a target knowledge graph;
step 3: and monitoring the change state of codes in the code warehouse in real time, and dynamically updating the target knowledge graph according to the change state data.
Preferably, extracting target knowledge elements from different types of data sources, integrating the target knowledge elements into a target meta-model, and establishing a target meta-data management system, wherein the method comprises the steps of;
extracting first knowledge elements from different types of data sources by utilizing an intelligent knowledge extraction function of a current metalevel low-code platform;
Monitoring the extraction process of the first knowledge element in real time, and responding when an abnormality is detected;
After the extracted first knowledge element data are cleaned, converting the data into formats and types supported by a target meta-model to obtain target knowledge elements;
Mapping the target knowledge elements to corresponding elements and relations in a target meta-model based on a set mapping rule;
Storing the mapping relation between the target knowledge elements and the corresponding elements and relations in the target meta-model to a target knowledge base;
Based on the data requirement, a target metadata management system is established to store the metadata of the current metalevel low-code platform.
Preferably, the data source types include structured types, semi-structured types, and unstructured types.
Preferably, the process of extracting the first knowledge element is monitored in real time, and in response to detecting an anomaly, the process includes:
Monitoring the first knowledge element extraction process in real time by using a target monitoring tool to obtain first performance index data;
performing state calibration on each first performance index data by comparing and analyzing the first performance index data with a set performance threshold;
Establishing an index state collection table according to a state calibration result of the first performance index data;
If only normal state indexes exist in the index state set table, judging that the current knowledge element extraction process is normal;
If the index state set table has abnormal state indexes or possible abnormal state indexes, judging that the current first knowledge element extraction process has abnormality;
When the first knowledge element extraction process is abnormal, adopting a corresponding abnormal processing strategy to perform abnormal processing.
Preferably, when there is an abnormality in the first knowledge element extraction process, a corresponding abnormality processing policy is adopted to perform abnormality processing, including:
If the index state set table has an abnormal state index, determining a first abnormal type based on the abnormal state index from a set type-abnormal processing mapping table, and extracting a first set abnormal processing rule matched with the current first abnormal type;
extracting an exception handling measure from the first set exception handling rule and taking the exception handling measure as a first measure;
If the index state set table has the possible abnormal state index, extracting historical performance data of the possible abnormal state index in a preset time period, and identifying to obtain a potential abnormal point by utilizing a machine learning algorithm;
Determining a first time window according to the timestamp of the potential abnormal point;
Extracting all log entries in a first time window from a log database, and screening the entries to obtain a first log entry;
Performing association analysis on the first log entry and potential abnormal points to obtain an abnormal association coefficient;
Marking a first log entry with an abnormal association coefficient greater than a set association threshold as a first high association entry;
Extracting the abnormality related features of the first high-association item, inputting the abnormality related features into a pre-established type recognition model, and outputting an abnormality type recognition result corresponding to the abnormality related features and taking the abnormality type recognition result as a second abnormality type;
Extracting a second set exception handling rule matched with the current second exception type from the set type-exception handling mapping table;
Extracting an exception handling measure from the second set exception handling rule and regarding the exception handling measure as a second measure;
When the index state set table has abnormal state indexes and no possible abnormal state indexes exist, immediately performing abnormal processing by taking the first measure as an abnormal processing strategy;
when the index state set table has possible abnormal state indexes and no abnormal state indexes, performing priority ranking on all acquired second measures according to a set priority evaluation mechanism, and then performing corresponding abnormal processing by taking the second measures as an abnormal processing strategy in sequence;
When the index state set table has an abnormal state index and a possible abnormal state index at the same time, immediately executing the acquired first measure to perform corresponding abnormal processing; if the second measure is consistent with the first measure, deleting the current second measure, sequencing the priority of the rest second measures according to a set priority evaluation mechanism, and then sequentially carrying out corresponding exception handling; if the second measure is consistent with the first measure, the acquired second measure is subjected to priority sorting according to a set priority evaluation mechanism, and then corresponding exception handling is performed in sequence.
Preferably, the calculation formula of the anomaly correlation coefficient is as follows:
; in the formula, An anomaly association coefficient represented as the i-th first target entry with the current potential anomaly point; A time interval represented as the time stamp of the ith first target entry and the occurrence time of the current potential outlier; text similarity expressed as description of the ith first target item and the current potential abnormal point; The total number of entities represented as the i-th first target entry; the number of the same entities represented as the ith first target item and the current potential abnormal point; an entity importance total value expressed as the same entity of the ith first target entry and the current potential outlier; the number of similar entities represented as the ith first target entry and the current potential outlier; similarity of the ith similar entity expressed as the ith first target entry to the current potential outlier; the entity importance of the jth similar entity represented as the ith first target entry and the current potential outlier.
Preferably, based on a preset graph database, implementing mapping construction of low-code development knowledge to obtain a target knowledge graph, including:
Selecting a set import mode to import the mapped target knowledge elements and the relation thereof into a preset map database;
Creating nodes and relations by using a target language provided by a preset graph database to obtain a target knowledge graph;
and visualizing the target knowledge graph by adopting a preset visualization tool.
Preferably, the method for monitoring the change state of the code in the code warehouse in real time and dynamically updating the target knowledge graph according to the change state data comprises the following steps:
Selecting a target monitoring mode from the monitoring modes based on analysis of screening requirement indexes to monitor changing conditions of a code warehouse, and obtaining code changing contents;
Extracting a first file associated with the target knowledge graph from the code change content by using a set matching rule;
Performing content analysis on the first file by adopting a set text processing mode to obtain target change information;
Determining a change type according to the target change information;
based on the change type, carrying out corresponding change on nodes, relations or attributes in the target knowledge graph to obtain a change knowledge graph;
and creating an index for the attribute of the frequently queried node and the attribute of the frequently queried relation in the changed knowledge graph.
Compared with the prior art, the application has the following beneficial effects:
Integrating the target knowledge elements into a meta model by extracting the target knowledge elements; based on a preset graph database, realizing the mapping construction of low-code development knowledge; the method monitors the change state of codes in the code warehouse in real time, dynamically updates the knowledge graph according to the change state data, can support rapid construction and updating of the software knowledge graph, is applied to various low-code development platforms, and improves development efficiency and software quality.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and drawings.
The technical scheme of the invention is further described in detail through the drawings and the embodiments.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:
Fig. 1 is a flowchart of a software knowledge graph intelligent construction method based on a metalevel low-code platform in an embodiment of the invention.
Detailed Description
The preferred embodiments of the present invention will be described below with reference to the accompanying drawings, it being understood that the preferred embodiments described herein are for illustration and explanation of the present invention only, and are not intended to limit the present invention.
The embodiment of the invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, which is shown in fig. 1 and comprises the following steps:
step 1: extracting target knowledge elements from different types of data sources, integrating the target knowledge elements into a target meta-model, and establishing a metadata management system;
Step 2: based on a preset graph database, realizing the mapping construction of the low-code development knowledge to obtain a target knowledge graph;
step 3: and monitoring the change state of codes in the code warehouse in real time, and dynamically updating the target knowledge graph according to the change state data.
In this embodiment, the low code platform refers to a development platform that allows a user to quickly build an application through a graphical interface and programming code; data source types include structured, semi-structured, and unstructured; the target knowledge elements comprise entities, relations and attribute knowledge elements, wherein the entity types comprise objects, concepts and events; relationship types include association, inheritance, dependency, aggregation, and so forth; the attribute type comprises basic attributes, associated attributes, characteristic attributes and dynamic attributes, wherein the basic attributes refer to basic information of an entity, such as an entity name and an identifier; the association attribute is used for describing association attributes of the entity and other entities, such as the category to which the entity belongs; the feature attributes are used to describe attributes of entity-specific features, such as entity size; dynamic properties refer to properties of an entity that change over time, such as status.
In this embodiment, the target meta-model refers to a pre-established model that covers core elements such as scene knowledge, demand knowledge, and template knowledge and specifies the structure and relationship between the core elements, where the scene knowledge refers to a knowledge set in a specific service scene, and is used to describe how an application should be constructed, configured, and used in a corresponding scene, including description of the scene, related service flows, required data models, and the like; the requirement knowledge refers to specific requirements of a user or a business party on an application, including functional requirements, performance requirements, interface requirements and the like; template knowledge refers to pre-defined reusable components, modules or frames in a low-code platform, including types, structures, attributes of templates, association relationships with other elements, and the like; the structure between core elements refers to the hierarchical structure, combination rules and constraints of the core elements, and the relationships include association, inheritance, dependence, aggregation and the like.
In this embodiment, the preset graph database is a predetermined graph database for knowledge graph construction, such as Neo4J, where the graph database refers to a data management system based on point and edge and using efficient storage and query graph data as design principles, and may be applied to the field of knowledge graphs, including Neo4J, janusGraph, galaxybase, tuGraph and so on.
In this embodiment, the target metadata management system is configured to store and manage various metadata in the metadata-level low-code platform, and is composed of a data storage module, a data management module, and a data utilization module, where the data storage module is configured to define a unified metadata format and standard, and store metadata; the data management module is used for realizing the functions of data input, data verification, version control and authority management; the data utilization module is used for realizing the functions of data query and analysis, data sharing and data visualization; code warehouses refer to a centralized or distributed storage space for storing, managing, tracking, and controlling code changes.
The beneficial effects of the technical scheme are as follows: integrating the target knowledge elements into a meta model by extracting the target knowledge elements; based on a preset graph database, realizing the mapping construction of low-code development knowledge; the method monitors the change state of codes in the code warehouse in real time, dynamically updates the knowledge graph according to the change state data, can support rapid construction and updating of the software knowledge graph, is applied to various low-code development platforms, and improves development efficiency and software quality.
The embodiment of the invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, which extracts target knowledge elements from different types of data sources, integrates the target knowledge elements into a target meta-model, and establishes a target meta-data management system, and comprises the following steps:
extracting first knowledge elements from different types of data sources by utilizing an intelligent knowledge extraction function of a current metalevel low-code platform;
Monitoring the extraction process of the first knowledge element in real time, and responding when an abnormality is detected;
After the extracted first knowledge element data are cleaned, converting the data into formats and types supported by a target meta-model to obtain target knowledge elements;
Mapping the target knowledge elements to corresponding elements and relations in a target meta-model based on a set mapping rule;
Storing the mapping relation between the target knowledge elements and the corresponding elements and relations in the target meta-model to a target knowledge base;
Based on the data requirement, a target metadata management system is established to store the metadata of the current metalevel low-code platform.
In this embodiment, the low code platform refers to a development platform that allows a user to quickly build an application through a graphical interface and programming code; the intelligent knowledge extraction function is used for extracting knowledge elements of entities, relations and attributes from different types of data sources; data source types include structured, semi-structured, and unstructured; the target knowledge elements comprise entities, relationships and attribute knowledge elements; the target meta-model is a pre-established model which covers core elements such as scene knowledge, demand knowledge and template knowledge and defines the structure and the relation among the core elements, wherein the scene knowledge is a knowledge set under a specific service scene and is used for describing how an application should be constructed, configured and used under a corresponding scene, and the scene knowledge comprises description of the scene, related service flows, required data models and the like; the requirement knowledge refers to specific requirements of a user or a business party on an application, including functional requirements, performance requirements, interface requirements and the like; template knowledge refers to pre-defined reusable components, modules or frames in a low-code platform, including types, structures, attributes of templates, association relationships with other elements, and the like; the structure among the core elements refers to the hierarchical structure, combination rule and constraint condition of the core elements, and the relation comprises association, inheritance, dependence, aggregation and the like; mapping the target knowledge elements to corresponding elements and relationships within the target meta-model is achieved through a mapping tool of the low-code platform; the data cleansing includes removing duplicate data, correcting erroneous data, and processing missing values.
In this embodiment, the target knowledge base is pre-established and is used for storing and managing the mapped target knowledge elements and the relationship thereof, providing operations such as inquiring, updating, deleting and the like of the target knowledge base, and supporting version control and authority management of data; data requirements refer to operational requirements for metadata, such as storage, verification, querying, visualization, and so forth; the target metadata management system is used for storing and managing various metadata in the metalevel low-code platform and consists of a data storage module, a data management module and a data utilization module, wherein the data storage module is used for defining unified metadata formats and standards and storing metadata; the data management module is used for realizing the functions of data input, data verification, version control and authority management; the data utilization module is used for realizing the functions of query and analysis, data sharing and data visualization.
The beneficial effects of the technical scheme are as follows: by utilizing the intelligent knowledge extraction function of the metalevel low-code platform to extract knowledge elements in different types of data sources, processing and mapping the knowledge elements, and establishing a target metadata management system, the data integration efficiency and accuracy can be remarkably improved, and the development and maintenance cost can be reduced.
The embodiment of the invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, which monitors the extraction process of a first knowledge element in real time and responds when an abnormality is detected, and comprises the following steps:
Monitoring the first knowledge element extraction process in real time by using a target monitoring tool to obtain first performance index data;
performing state calibration on each first performance index data by comparing and analyzing the first performance index data with a set performance threshold;
Establishing an index state collection table according to a state calibration result of the first performance index data;
If only normal state indexes exist in the index state set table, judging that the current knowledge element extraction process is normal;
If the index state set table has abnormal state indexes or possible abnormal state indexes, judging that the current first knowledge element extraction process has abnormality;
When the first knowledge element extraction process is abnormal, adopting a corresponding abnormal processing strategy to perform abnormal processing.
In this embodiment, the target monitoring tool is a monitoring tool for monitoring the first knowledge element extraction process in real time, for example, promethaus open source monitoring; the first performance index data includes response time, throughput, resource utilization, error rate, and the like.
In this embodiment, the index state set table is established as follows:
(1) Comparing the first performance index data with a set performance threshold;
(2) If the first performance index data is not greater than the set performance threshold and does not exceed the corresponding preset index range, marking the current first performance index data as a normal state index;
(3) If the first performance index data is not greater than the set performance threshold and exceeds the corresponding preset index range, marking the current first performance index data as a possible abnormal state index;
(4) If the first performance index data is larger than the set performance threshold, marking the current first performance index data as an abnormal state index;
(5) And collecting the first performance index data after the state marking to obtain an index state collection table.
In this embodiment, the set performance threshold is a value obtained by adding the average value and the standard deviation obtained by calculating the historical performance data over a period of time; the preset index range is composed of a preset index upper limit and a preset index lower limit, wherein the preset index upper limit is a value accounting for 95% of the set performance threshold, and the preset index lower limit is a value accounting for 15% of the set performance threshold; the state calibration result comprises three types of normal, abnormal and possible abnormal; the index state set table is composed of normal state indexes, abnormal state indexes or possible abnormal state indexes.
In this embodiment, for example, there is a response time E1 of 60s, a set performance threshold E0 of 120s, and a preset difference range of the response time ofI.e.Where s is a time unit, expressed in seconds;
and if the response time E1 is smaller than the set performance threshold E0 and the response time E1 does not exceed the preset index range, marking the current response time E1 as a normal state index.
The beneficial effects of the technical scheme are as follows: the method has the advantages that the target monitoring tool is used for monitoring the knowledge element process in real time, and the performance state and potential problems are managed through the threshold comparison, judgment and response mechanism, so that the resource utilization efficiency can be improved, the problem response time can be shortened, and the processing efficiency can be improved.
The embodiment of the invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, which adopts a corresponding exception handling strategy to carry out exception handling when the first knowledge element extraction process is abnormal, and comprises the following steps:
If the index state set table has an abnormal state index, determining a first abnormal type based on the abnormal state index from a set type-abnormal processing mapping table, and extracting a first set abnormal processing rule matched with the current first abnormal type;
extracting an exception handling measure from the first set exception handling rule and taking the exception handling measure as a first measure;
If the index state set table has the possible abnormal state index, extracting historical performance data of the possible abnormal state index in a preset time period, and identifying to obtain a potential abnormal point by utilizing a machine learning algorithm;
Determining a first time window according to the timestamp of the potential abnormal point;
Extracting all log entries in a first time window from a log database, and screening the entries to obtain a first log entry;
Performing association analysis on the first log entry and potential abnormal points to obtain an abnormal association coefficient;
Marking a first log entry with an abnormal association coefficient greater than a set association threshold as a first high association entry;
Extracting the abnormality related features of the first high-association item, inputting the abnormality related features into a pre-established type recognition model, and outputting an abnormality type recognition result corresponding to the abnormality related features and taking the abnormality type recognition result as a second abnormality type;
Extracting a second set exception handling rule matched with the current second exception type from the set type-exception handling mapping table;
Extracting an exception handling measure from the second set exception handling rule and regarding the exception handling measure as a second measure;
When the index state set table has abnormal state indexes and no possible abnormal state indexes, immediately performing abnormal processing by taking the first measure as an abnormal processing strategy;
When the index state set table has abnormal state indexes and no possible abnormal state indexes exist, immediately performing abnormal processing by taking the first measure as an abnormal processing strategy;
when the index state set table has possible abnormal state indexes and no abnormal state indexes, performing priority ranking on all acquired second measures according to a set priority evaluation mechanism, and then performing corresponding abnormal processing by taking the second measures as an abnormal processing strategy in sequence;
When the index state set table has an abnormal state index and a possible abnormal state index at the same time, immediately executing the acquired first measure to perform corresponding abnormal processing; if the second measure is consistent with the first measure, deleting the current second measure, sequencing the priority of the rest second measures according to a set priority evaluation mechanism, and then sequentially carrying out corresponding exception handling; if the second measure is consistent with the first measure, the acquired second measure is subjected to priority sorting according to a set priority evaluation mechanism, and then corresponding exception handling is performed in sequence.
In the embodiment, a set type-exception handling mapping table is pre-established based on exception data and a corresponding successful handling strategy in a historical knowledge extraction process, and is composed of an exception type, an exception type description, a performance index and a set exception handling rule, wherein the exception type comprises data exception, performance exception, system error and logic error; the set exception handling rules are composed of exception identification types and exception handling measures, wherein the exception handling measures comprise data rollback, resource allocation adjustment, extraction rule optimization, mapping logic optimization, logging, personnel notification and the like.
In this embodiment, the first exception type is a type selected from a set type-exception handling mapping table according to an exception status index; the first set exception handling rule refers to a set exception handling rule matched with the first exception type; the first measure refers to an exception handling measure extracted from a first set exception handling rule.
In this embodiment, the preset time period is preset; the potential abnormal points are identification results output after the historical performance data are input into an abnormal identification model which is established in advance based on a machine learning algorithm, wherein the abnormal identification model is a model obtained by training an isolated forest model by utilizing the pre-acquired historical data related to the performance index; the first time window is a time period determined based on the time stamp of the potential outlier, such as the time stamp of the existence of the potential outlier x1, i.e. the occurrence time is T0, and the starting time of the first time window isThe end time isWherein 0.5Expressed as half an hour, h is the time unit expressed as hours.
In this embodiment, the log database is formed by log data of each component such as a system, an application program, a database, and the like; the log entries refer to all entries in a first time window screened from a log database, and the entries refer to record information automatically generated in the running process of a computer system and generally comprise the occurrence time of specific events, the types of the events, related components or modules, error codes, detailed description information and the like; the first log entry refers to an entry selected from the obtained target entries by taking a specific keyword as a selection condition, wherein the keyword includes a component name, warning information, and the like related to the performance index.
In this embodiment, the anomaly correlation coefficients are used to quantify the degree of correlation between log entries and potential anomaly points; setting a correlation threshold value refers to the average value of correlation coefficients of different historical log entries of preset quantity and known abnormal points and the average value of peaks; the first high-association entry is a first log entry with an abnormal association coefficient larger than a set association threshold; exception-related features include error codes, exception information, performance index changes associated with exceptions, such as extended response time, increased resource consumption, specific behavior of system components, and the like.
In this embodiment, the type recognition model is a model obtained by training a neural network by using, as training data, historical anomaly characteristics extracted from a preset amount of normal log entries and anomaly log entries, for recognizing anomaly types; the second abnormality type refers to an abnormality related characteristic of the first high-association item is input into a type recognition model, and an output abnormality type recognition result is output; the second set exception handling rule refers to a preset set exception handling rule matched with the second exception type in the set type-exception handling mapping table; the second measure is an exception handling measure extracted from a second set exception handling rule; the priority evaluation mechanism is set to acquire a priority coefficient of the second measure and prioritize the second measure according to the order from large to small, wherein the priority coefficient is a value obtained by multiplying an absolute difference value between a corresponding possible state index of the current second measure and a set performance threshold value and an index importance weight of the corresponding possible state index, wherein the index importance weight is a weight value calculated by using a hierarchical analysis method to target the extraction speed of a knowledge element and determining the relative importance of the index by comparing every two.
The beneficial effects of the technical scheme are as follows: when the knowledge element extraction process is abnormal, the corresponding abnormality processing strategy is criticized for performing abnormality processing, so that the construction efficiency of the knowledge graph can be improved, the construction accuracy of the knowledge graph is ensured, and the reliability of the knowledge graph is further enhanced.
The embodiment of the invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, wherein the calculation formula of an abnormal association coefficient is as follows:
; in the formula, An anomaly association coefficient represented as the i-th first target entry with the current potential anomaly point; A time interval represented as the time stamp of the ith first target entry and the occurrence time of the current potential outlier; text similarity expressed as description of the ith first target item and the current potential abnormal point; The total number of entities represented as the i-th first target entry; the number of the same entities represented as the ith first target item and the current potential abnormal point; an entity importance total value expressed as the same entity of the ith first target entry and the current potential outlier; the number of similar entities represented as the ith first target entry and the current potential outlier; similarity of the ith similar entity expressed as the ith first target entry to the current potential outlier; the entity importance of the jth similar entity represented as the ith first target entry and the current potential outlier.
In this embodiment, the anomaly correlation coefficients are used to quantify the degree of correlation between log entries and potential anomaly points; the timestamp of the first target entry refers to the point in time when the first target entry was generated; the potential outlier description refers to a keyword and an error code; the text similarity is a numerical value which is calculated by utilizing a TF-IDF algorithm and used for quantifying the text similarity degree of the first target item and the current potential abnormal point description; the entity refers to a service name, a file name, an ip address, etc.; the entity importance refers to a numerical value for quantifying the entity importance, and is determined by calculating the frequency of occurrence of the entity in all first target items; the similarity of similar entities is calculated by using a cosine similarity algorithm.
The beneficial effects of the technical scheme are as follows: and selecting a log entry highly associated with the potential abnormal point by calculating the abnormal association degree, providing data support for identifying the current abnormal type, further facilitating the adoption of an accurate abnormal processing strategy for carrying out abnormal processing, and improving the knowledge element extraction efficiency.
The embodiment of the invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, which is used for realizing the graph construction of low-code development knowledge based on a preset graph database to obtain a target knowledge graph, and comprises the following steps:
Selecting a set import mode to import the mapped target knowledge elements and the relation thereof into a preset map database;
Creating nodes and relations by using a target language provided by a preset graph database to obtain a target knowledge graph;
and visualizing the target knowledge graph by adopting a preset visualization tool.
In this embodiment, the set import mode is used to import the mapped target knowledge elements and their relationships into the preset map database, for example, when the preset map database is Neo4j map database, the import tool of the corresponding set import mode for importing large-scale data is Neo4j, the corresponding set import mode for importing small-scale or incremental data is Cypher language, where the size of the data scale is determined by the amount of imported data, and when the amount of imported data exceeds thirty-thousand pieces, or the size of the imported data file exceeds five hundred megabytes, the large-scale is determined; a small scale is determined when the amount of imported data does not exceed thirty-thousand pieces, or the imported data file size does not exceed five hundred megabytes.
In this embodiment, the preset graph database is a predetermined graph database for knowledge graph construction, for example, neo4J graph database refers to a high-performance NoSQL graph database storing data according to the characteristics of a graph model; the graph database is a data management system which takes points and edges as basic storage units and takes efficient storage and query graph data as design principles, and can be applied to the field of knowledge graphs, including Neo4J, janusGraph, galaxybase, tuGraph and the like; the target language refers to a language used for realizing creation, modification and deletion of nodes and relations in a preset graph database, for example, a Cypher language in a Neo4J graph database; the target knowledge graph is a knowledge graph developed by low codes gradually constructed by creating and connecting nodes in a preset graph database; the preset visualization tool is used for realizing the visualization of the target knowledge graph, when the preset graph database is internally provided with the visualization tool, the corresponding built-in visualization tool is regarded as the preset visualization tool, for example, neo4J Browser in the Neo4J graph database is used for visualizing the target knowledge graph by using a third party visualization tool compatible with the current preset graph database as the preset visualization tool when the preset graph database is not internally provided with the visualization tool.
The beneficial effects of the technical scheme are as follows: the mapped target knowledge elements and the relation thereof are imported into a preset graph database to form a knowledge graph and visualized, so that the rapid construction of the software knowledge graph can be supported, and the development efficiency and the software quality are improved.
The embodiment of the invention provides a software knowledge graph intelligent construction method based on a metalevel low-code platform, which monitors the change state of codes in a code warehouse in real time and realizes the dynamic update of a target knowledge graph according to change state data, and comprises the following steps:
Selecting a target monitoring mode from the monitoring modes based on analysis of screening requirement indexes to monitor changing conditions of a code warehouse, and obtaining code changing contents;
Extracting a first file associated with the target knowledge graph from the code change content by using a set matching rule;
Performing content analysis on the first file by adopting a set text processing mode to obtain target change information;
Determining a change type according to the target change information;
based on the change type, carrying out corresponding change on nodes, relations or attributes in the target knowledge graph to obtain a change knowledge graph;
and creating an index for the attribute of the frequently queried node and the attribute of the frequently queried relation in the changed knowledge graph.
In this embodiment, the monitoring manner is used to monitor the change condition of the code repository, including Webhook, timing polling, etc., where the code repository refers to a centralized or distributed storage space for storing, managing, tracking, and controlling code changes; the target monitoring mode is a monitoring mode with the highest monitoring adaptation coefficient selected from monitoring modes supporting the current code warehouse, wherein the monitoring adaptation coefficient is a coefficient obtained by averaging index evaluation values obtained by evaluating the monitoring of the current code warehouse based on screening requirement indexes; screening requirement indexes including real-time requirements, resource consumption requirements and safety requirements; setting a matching rule to be text path mode matching; the code change content refers to a file list and content which are newly added, modified or deleted; the change state data refers to code change contents.
In the embodiment, setting a matching rule refers to traversing acquired code change contents by taking a file name and keywords in file contents as matching conditions, wherein the keywords refer to business logic keywords of software, and the keywords comprise class names, method names, variable names and the like; the first file is a file which is extracted from the code change content by using a set matching rule and is associated with the target knowledge graph; setting a text processing mode to be predetermined, wherein the text processing mode comprises a regular expression and a corresponding parsing library of a special format, and the special format comprises Java, JSON, XML and the like; the target change information is information obtained by carrying out content analysis on the first file and comprises a new function, a modified field, a class name and the like; the change types include node addition, relationship modification, attribute update, and the like; the change knowledge graph is obtained by correspondingly changing nodes, relations or attributes in the target knowledge graph according to the change type, wherein the nodes refer to entities or concepts; relationships refer to relationships between different entities, such as inheritance relationships, dependency relationships, causal relationships, and the like; attributes refer to specific features describing an entity, such as the name, identifier, category to which the entity belongs, and so forth; the frequent query node refers to a node with the query times exceeding the set query times; frequent query relationships refer to relationships in which the number of queries exceeds a set number of queries; the set query times refer to the query peak values of the node attributes or the relationship attributes in a preset time period.
The beneficial effects of the technical scheme are as follows: the dynamic updating of the target knowledge graph is effectively realized by monitoring the change state of codes in the code warehouse in real time and analyzing the change state data, so that the timeliness of the knowledge graph is further ensured.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (8)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411449597.8A CN118964682B (en) | 2024-10-17 | 2024-10-17 | Intelligent software knowledge graph construction method based on metalevel low-code platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411449597.8A CN118964682B (en) | 2024-10-17 | 2024-10-17 | Intelligent software knowledge graph construction method based on metalevel low-code platform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118964682A true CN118964682A (en) | 2024-11-15 |
| CN118964682B CN118964682B (en) | 2024-12-20 |
Family
ID=93402670
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411449597.8A Active CN118964682B (en) | 2024-10-17 | 2024-10-17 | Intelligent software knowledge graph construction method based on metalevel low-code platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118964682B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119494395A (en) * | 2025-01-20 | 2025-02-21 | 东北大学 | A method for constructing a dynamic knowledge graph for sintering |
| CN119740649A (en) * | 2025-03-03 | 2025-04-01 | 北京大学 | Software knowledge graph construction method and system for low-code template recommendation |
| CN119862174A (en) * | 2025-03-24 | 2025-04-22 | 中船奥蓝托无锡软件技术有限公司 | Low-code database merging method based on incremental meta-model atlas |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160055409A1 (en) * | 2014-08-19 | 2016-02-25 | Qualcomm Incorporated | Knowledge-graph biased classification for data |
| CN114035783A (en) * | 2021-11-10 | 2022-02-11 | 成都淞幸科技有限责任公司 | Software code knowledge graph construction method and tool |
| CN114493255A (en) * | 2022-01-25 | 2022-05-13 | 平安国际智慧城市科技股份有限公司 | Enterprise anomaly monitoring method and related equipment based on knowledge graph |
| CN115543402A (en) * | 2022-11-21 | 2022-12-30 | 北京大学 | A Method for Incremental Update of Software Knowledge Graph Based on Code Submission |
| CN116595191A (en) * | 2023-05-17 | 2023-08-15 | 北京无代码科技有限公司 | Construction method and device of interactive low-code knowledge graph |
-
2024
- 2024-10-17 CN CN202411449597.8A patent/CN118964682B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160055409A1 (en) * | 2014-08-19 | 2016-02-25 | Qualcomm Incorporated | Knowledge-graph biased classification for data |
| CN114035783A (en) * | 2021-11-10 | 2022-02-11 | 成都淞幸科技有限责任公司 | Software code knowledge graph construction method and tool |
| CN114493255A (en) * | 2022-01-25 | 2022-05-13 | 平安国际智慧城市科技股份有限公司 | Enterprise anomaly monitoring method and related equipment based on knowledge graph |
| CN115543402A (en) * | 2022-11-21 | 2022-12-30 | 北京大学 | A Method for Incremental Update of Software Knowledge Graph Based on Code Submission |
| CN116595191A (en) * | 2023-05-17 | 2023-08-15 | 北京无代码科技有限公司 | Construction method and device of interactive low-code knowledge graph |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119494395A (en) * | 2025-01-20 | 2025-02-21 | 东北大学 | A method for constructing a dynamic knowledge graph for sintering |
| CN119740649A (en) * | 2025-03-03 | 2025-04-01 | 北京大学 | Software knowledge graph construction method and system for low-code template recommendation |
| CN119862174A (en) * | 2025-03-24 | 2025-04-22 | 中船奥蓝托无锡软件技术有限公司 | Low-code database merging method based on incremental meta-model atlas |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118964682B (en) | 2024-12-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN118964682B (en) | Intelligent software knowledge graph construction method based on metalevel low-code platform | |
| CN108197132B (en) | Graph database-based electric power asset portrait construction method and device | |
| US9348943B2 (en) | Method for analyzing time series activity streams and devices thereof | |
| CN116894152A (en) | Multisource data investigation and real-time analysis method | |
| US20170109668A1 (en) | Model for Linking Between Nonconsecutively Performed Steps in a Business Process | |
| CN112711591A (en) | Data blood margin determination method and device based on field level of knowledge graph | |
| US20170109667A1 (en) | Automaton-Based Identification of Executions of a Business Process | |
| CN117971606B (en) | Log management system and method based on elastic search | |
| EP2633490A1 (en) | Providing information management | |
| CN117931913A (en) | A multi-agent data space interoperability method for multi-source heterogeneous data | |
| CN115221337A (en) | Data weaving processing method, device, electronic device and readable storage medium | |
| CN119336831A (en) | A method for intelligent search of power data based on resource map | |
| CN119557845A (en) | A method, device and equipment for multi-source heterogeneous data fusion processing and intelligent analysis | |
| CN119494463A (en) | A method for monitoring urban operation indicators based on multi-source heterogeneity | |
| US11308104B2 (en) | Knowledge graph-based lineage tracking | |
| CN113673889A (en) | Intelligent data asset identification method | |
| CN117171243A (en) | Data standardized management method and system | |
| CN118898234B (en) | A system and method for automatically labeling government big data | |
| CN119515352A (en) | A television intelligent maintenance decision-making method, system and computer program | |
| CN117311777A (en) | Automatic operation and maintenance platform and method | |
| CN117312303A (en) | Automatic data asset checking method, device, electronic equipment and medium | |
| CN117435792A (en) | Distributed data braiding processing architecture | |
| CN114139979A (en) | Service platform for specific research and development mechanism | |
| CN119904154B (en) | Artificial intelligence modeling method, device, equipment and medium based on reporting system | |
| CN112214660A (en) | Industrial Information Identification and Retrieval System |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |