CN115033746A - Ship navigation accident cause analysis method based on root-tying theory and incident map - Google Patents
Ship navigation accident cause analysis method based on root-tying theory and incident map Download PDFInfo
- Publication number
- CN115033746A CN115033746A CN202210608600.0A CN202210608600A CN115033746A CN 115033746 A CN115033746 A CN 115033746A CN 202210608600 A CN202210608600 A CN 202210608600A CN 115033746 A CN115033746 A CN 115033746A
- Authority
- CN
- China
- Prior art keywords
- ship navigation
- accident
- event
- theory
- navigation accident
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/25—Integrating or interfacing systems involving database management systems
- G06F16/254—Extract, transform and load [ETL] procedures, e.g. ETL data flows in data warehouses
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A ship navigation accident cause analysis method based on a rooting theory and a case map comprises the steps of firstly obtaining various ship navigation accident reports, and constructing a ship navigation accident field ontology core concept through a self-defined core concept based on the rooting theory; classifying, labeling and extracting the event entities according to the constructed ontology core concept; labeling and extracting the event relation based on the event extraction result; importing the obtained results of the extraction of the ship navigation accident event and the extraction of the event relation into a Neo4j database to complete the construction of a ship navigation accident event map; and carrying out cause analysis on the ship navigation accident based on the constructed ship navigation accident case map. The method is based on the root-tying theory and combined with the characteristics of the field of ship navigation accidents, the factors of the ship navigation accidents are extracted, the model of the factors of the ship navigation accidents is constructed, the factors of the ship navigation accidents are systematically analyzed, and the method has theoretical and practical significance for ship navigation safety management.
Description
Technical Field
The invention belongs to the technical field of a matter graph, relates to a root theory, a matter graph and an accident cause analysis cross technology, and particularly relates to a method for realizing analysis of a cause of a ship navigation accident by constructing the matter graph through the root theory.
Background
Shipping is one of the important transportation means for ensuring international trade and global economic prosperity, and water transportation accounts for about 90% of world trade volume by 2020 according to statistics. Shipping activities are becoming more frequent, thus also leading to a high incidence of accidents. In order to prevent accidents or reduce the influence of accidents to the maximum extent, accident-causing elements need to be found out as accurately and completely as possible and corresponding safety precautionary measures need to be taken. Therefore, analyzing the cause of the ship navigation accident is an important means for improving the safety of the ship navigation. The existing cause analysis about ship navigation accidents mostly adopts a linear accident chain cause model or an infectious disease cause model, the result of researching the ship accidents from a systematic visual angle is less, careful and solid large sample research is lacked, and the complex influence of the navigation accidents is difficult to capture.
Disclosure of Invention
The invention aims to provide a ship navigation accident cause analysis method based on a rooting theory and a physical map, aiming at the defects in the prior art.
Aiming at the current situations of complex causes, strong emergencies and high danger degree of ship accidents, the invention applies a research method combining a rooting theory and a matter map, constructs a navigation accident corpus based on the rooting theory by analyzing a ship navigation accident survey report and combining the field characteristics of the ship navigation accidents, and constructs the ship navigation accident matter map based on a deep learning method on the basis, thereby analyzing each cause of the ship navigation accidents.
In order to achieve the purpose, the invention is realized by adopting the following technical scheme.
A ship navigation accident cause analysis method based on a rooting theory and a matter map comprises the following steps:
1. acquiring various ship navigation accident investigation reports;
2. constructing a body core concept of the ship navigation accident field based on a root theory by combining the characteristics of the ship navigation accident and the reuse of the existing bodies of other accident fields for the ship navigation accident investigation report obtained in the step 1;
3. classifying ship navigation accident event entities in the ship navigation accident survey report obtained in the step 1 based on the ship navigation accident body core concept constructed in the step 2, labeling the event entities based on the body core concept to establish a set of event map corpus suitable for the field of ship navigation accidents, labeling the event entities to extract events in a supervised learning mode, and completing construction of the events in the ship navigation accident event map;
4. based on the event entity extraction result in the step 3, adopting a deep learning tool to label the event relation of the ship navigation accident in the ship navigation accident investigation report obtained in the step 1, and then adopting a Pulse Coupled Neural Network (PCNN) method to semi-automatically extract the labeled event relation so as to complete the construction of the event relation in the ship navigation case map;
5. importing the results of the ship navigation accident event extraction and the event relation extraction obtained in the steps 3 and 4 into a Neo4j database to complete the construction of a ship navigation accident event map;
6. and (5) carrying out cause analysis on the ship navigation accident based on the ship navigation accident case map constructed in the step (5).
Preferably, in step 2, the constructing of the core concept refers to performing openness coding, main axis coding and selective coding on the data in the survey report of the ship navigation accident in step 1, then extracting the core concept of the ship navigation accident, and finally forming the core concept of the ship navigation accident body.
Preferably, the openness code is used for coding the original data of the survey report of the ship navigation accident in the step 1 line by line and sentence by sentence, gradually conceptualizing and categorizing the original data by continuously comparing, and eliminating the initial concept with the frequency below three times, so as to extract the initial category; the main shaft coding further refines the initial category formed by the open coding, divides different types of the initial category, classifies categories with similar structure, type and logic relationship into one category, and further distinguishes the main category and the corresponding sub-category; the selective coding refers to the steps of sorting the main scope, comparing human causes (Men), equipment causes (Machine), operation causes (Media) and Management causes (Management) of a 4M theoretical model, combining the characteristics of ship navigation accidents, selecting the core scope, systematically connecting the core scope with other scopes, verifying the relation between the core scope and the other scopes, and supplementing and completing the scope which is conceptualized and is not completely developed. The core concept extraction means that firstly, event participation roles in the SEM model are used as an upper body layer through events and scenes in the ABC body model, and types, places and events in the SEM model are reused as a primary concept; and then, referring to a plurality of accident related field ontologies such as chemical accidents, building accidents, emergency events and the like, extracting and generalizing related core concepts, and finally combining a selective code formed by a root theory method to complete the construction of the core concept of the ship navigation accident ontology.
Further preferably, the main category refers to safety management, perception error, illegal operation, decision error, environmental factor, ship equipment factor, personnel factor, improper crew arrangement and insufficient navigation plan, and the core category refers to equipment factor, environmental factor, management factor, human factor and emergency response factor.
Preferably, the supervised learning mode in the step 3 is to divide the case map corpus into a training set, a test set and a verification set according to the ratio of 8:1:1, vectorize the corpus by using a Word2vec tool, and input a vector into a two-way long-short time memory neural network and a conditional random field BilSTM-CRF model to perform event extraction training.
Further preferably, in the PCNN method described in step 4, the survey report text of the ship navigation accident obtained in step (1) is divided into three sections for convolution, then the maximum pooling is performed in sections, and finally the semi-automatic extraction of the event relation is completed through a classifier.
Further preferably, in the analysis of the cause of the ship navigation accident in the step 6, the key element path of the cause of the accident is mined from the ship navigation event map obtained in the step (5), then the accident conduction paths changing with time are analyzed in different time periods, the key conduction paths of different types of accidents are analyzed, finally, the central maps of different types of key accidents are extracted, and the cause analysis of specific accidents is carried out.
Further preferably, the mining of the key element path for the cause of the mining accident further comprises the following steps:
a) the predefined critical value represents the minimum number of accident-cause-effect pairs of significant element path connections;
b) defining a seed accident set, traversing the reason nodes from a given seed accident, performing breadth traversal and depth traversal on the case graph to form a causal path tuple list of the seed nodes
c) Calculating the causality of each cause node in the traversal process, checking whether the causality is larger than a given critical value, if so, retaining, and otherwise, pruning.
d) And determining the element path weight of the mined accident element path by a semi-supervised learning PU method, namely respectively taking the seed accident and the candidate entity as positive example data and non-label data, confirming a reliable negative example set from the non-label data, and then establishing a classifier by adopting SVM iteration so as to automatically learn the element path weight and finally obtain the element path ordered according to importance.
The method analyzes and researches the cause of the ship navigation accident by constructing a ship navigation accident case map. Aiming at various ship navigation accident survey reports, an ontology core concept is constructed through a root-tying theory so as to realize event extraction and event relation extraction, a case map capable of describing the cause of a ship navigation accident is established, and the cause of the ship navigation accident is analyzed based on the case map.
Drawings
FIG. 1 is a flow chart of a ship navigation accident cause analysis method based on rooting theory and a physical map.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, the specific contents and steps of the method for analyzing the cause of a ship navigation accident based on the rooting theory and the physiological map are as follows:
(1) acquiring various ship navigation accident investigation reports;
(2) constructing a ship navigation accident field body core concept based on a root theory by combining the characteristics of the ship navigation accident and the reuse of the existing body of other accident fields for the ship navigation accident investigation report obtained in the step (1); and (3) constructing a core concept, namely performing openness coding, main shaft coding and selective coding on the data in the survey report of the ship navigation accident in the step (1). The openness coding is used for coding the original data of the ship navigation accident survey report in the step (1) line by line and sentence by sentence, gradually conceptualizing and categorizing the original data through continuous comparison, and eliminating the initial concept with the frequency below three times, so that the initial category is extracted; the main shaft coding further refines the initial category formed by the open coding, divides different types of the initial category, classifies categories with similar structure, type and logic relationship into one category, and further distinguishes the main category and the corresponding sub-category; the selective coding refers to the steps of sorting the main scope, comparing human causes (Men), equipment causes (Machine), operation causes (Media) and Management causes (Management) of a 4M theoretical model, combining the characteristics of ship navigation accidents, selecting the core scope, systematically connecting the core scope with other scopes, verifying the relation between the core scope and the other scopes, and supplementing and completing the scope which is conceptualized and is not completely developed. The main category refers to safety management, perception error, illegal operation, decision error, environmental factors, ship equipment factors, personnel factors, improper arrangement of crews and insufficient navigation plan, and the core category refers to equipment factors, environmental factors, management factors, human factors and emergency response factors; the core concept extraction means that events and situations in an ABC body model and event participation roles in an SEM model are used as an upper body layer, types, places and events in the SEM model are reused as a primary concept, then a plurality of accident related field bodies such as chemical accidents, building accidents and emergent events are referred, the related core concepts are extracted and generalized, and finally the construction of the core concept of the ship navigation accident body is completed by combining selective codes formed by a root-tying theoretical method.
(3) Classifying ship navigation accident event entities in the ship navigation accident investigation report acquired in the step (1) based on the ship navigation accident body core concept constructed in the step (2), labeling the event entities based on the body core concept to establish a set of event map corpus suitable for the ship navigation accident field, and labeling the event entities to perform event extraction in a supervised learning mode; the supervised learning mode is characterized in that a matter atlas corpus is divided into a training set, a testing set and a verification set according to the ratio of 8:1:1, a Word2vec tool is used for vectorizing the matter atlas corpus, and then vectors are input into a bidirectional long-short time memory neural network and a conditional random field BilSTM-CRF model for event extraction training to complete the construction of events in the ship navigation case atlas;
(4) based on the event extraction result in the step (3), marking the event relation of the ship navigation accident in the ship navigation accident investigation report obtained in the step (1) by adopting a deep learning tool, and then semi-automatically extracting the marked event relation by adopting a Pulse Coupled Neural Network (PCNN) method; the PCNN method comprises the steps of dividing a ship navigation accident survey report text obtained in the step (1) into three sections for convolution, then carrying out sectional maximum pooling, and finally completing semi-automatic extraction of event relations through a classifier to complete construction of event relations in a ship navigation case map;
(5) importing the results of the ship navigation accident event extraction and the event relation extraction obtained in the steps (3) and (4) into a Neo4j database to complete the construction of a ship navigation accident case map;
(6) and (4) mining key element paths of accident cause based on the ship navigation accident case map constructed in the step (5), analyzing accident conduction paths which change along with time in different time periods, analyzing key conduction paths of different types of accidents, and finally extracting central maps of different types of key accidents to perform cause analysis of specific accidents.
The mining of the accident cause key element path in the step (6) further comprises the following steps:
a) the predefined critical value represents the minimum number of accident-cause-effect pairs of significant element path connections;
b) defining a seed accident set, traversing the reason nodes from a given seed accident, performing breadth traversal and depth traversal on the case graph to form a causal path tuple list of the seed nodes
c) Calculating the causality of each cause node in the traversal process, checking whether the causality is larger than a given critical value, if so, retaining, and otherwise, pruning.
d) And determining the element path weight of the mined accident element path by a semi-supervised learning PU method, namely respectively taking the seed accident and the candidate entity as positive example data and non-label data, confirming a reliable negative example set from the non-label data, and then establishing a classifier by adopting SVM iteration so as to automatically learn the element path weight and finally obtain the element path ordered according to importance.
Claims (8)
1. A ship navigation accident cause analysis method based on a rooting theory and a affair map is characterized by comprising the following steps:
(1) acquiring various ship navigation accident investigation reports;
(2) constructing a ship navigation accident field body core concept based on a root theory by combining the characteristics of the ship navigation accident and the reuse of the existing body of other accident fields for the ship navigation accident investigation report obtained in the step (1);
(3) classifying ship navigation accident event entities in the ship navigation accident investigation report acquired in the step (1) based on the ship navigation accident body core concept constructed in the step (2), labeling the event entities based on the body core concept to establish a set of event map corpus suitable for the ship navigation accident field, and labeling the event entities to perform event extraction in a supervised learning mode;
(4) based on the event entity extraction result in the step (3), adopting a deep learning tool to label the event relation of the ship navigation accident in the ship navigation accident investigation report obtained in the step (1), and then adopting a Pulse Coupled Neural Network (PCNN) method to semi-automatically extract the labeled event relation;
(5) importing the results of the ship navigation accident event extraction and the event relation extraction obtained in the steps (3) and (4) into a Neo4j database to complete the construction of a ship navigation accident case map;
(6) and (5) carrying out cause analysis on the ship navigation accident based on the ship navigation accident case map constructed in the step (5).
2. The ship navigation accident cause analysis method based on rooting theory and event graph according to claim 1, characterized in that in step (2), the specific content and method for constructing the ontology core concept is to perform openness coding, main shaft coding and selective coding on the data in the ship navigation accident survey report in step (1), and then extract the ship navigation accident core concept to finally form the ship navigation accident ontology core concept.
3. The ship navigation accident cause analysis method based on the rooting theory and the matter graph according to the claim 1, characterized in that in the step (3), the supervised learning mode is that the matter graph corpus is divided into a training set, a testing set and a verification set according to the ratio of 8:1:1, a Word2vec tool is used for vectorizing the matter graph corpus, then the vector is input into a two-way long-short time memory neural network and a conditional random field BilSTM-CRF model for event extraction training, and finally the construction of the event in the ship navigation accident matter graph is completed.
4. The method for analyzing the cause of the ship navigation accident based on the rooting theory and the event graph according to claim 1, wherein in the step (4), the specific content and the steps of the pulse coupled neural network PCNN method are that the ship navigation accident survey report text obtained in the step (1) is divided into three sections to be convoluted, then the subsection maximal pooling is carried out, the semi-automatic extraction of the event relation is completed through a classifier, and finally the construction of the event relation in the ship navigation case graph is completed.
5. The ship navigation accident cause analysis method based on rooting theory and event graph according to claim 1, characterized in that in step (6), the ship navigation accident cause analysis includes digging an accident cause key element path from the ship navigation event graph obtained in step (5), analyzing accident conduction paths varying with time in different time periods, analyzing key conduction paths of different types of accidents, and finally extracting central graphs of different types of key accidents to perform cause analysis of specific accidents.
6. The ship navigation accident cause analysis method based on the rooting theory and the event graph according to claim 5, wherein the specific content and the method steps of excavating the key element path of the accident cause are as follows:
a) the predefined critical value represents the minimum number of accident-cause-effect pairs of significant element path connections;
b) defining a seed accident set, traversing reason nodes from a given seed accident, and performing breadth traversal and depth traversal on the case graph to form a causal path tuple list of seed nodes;
c) calculating the causality of each cause node in the traversal process, checking whether the causality is larger than a given critical value, if so, reserving, and otherwise, pruning;
d) and determining the element path weight of the mined accident element path by a semi-supervised learning PU method, namely respectively taking the seed accident and the candidate entity as positive example data and non-label data, confirming a reliable negative example set from the non-label data, and then establishing a classifier by adopting SVM iteration so as to automatically learn the element path weight and finally obtain the element path ordered according to importance.
7. The ship navigation accident cause analysis method based on the rooting theory and the affair map according to claim 2, characterized in that the specific content and method steps of the openness code are to code the original data of the ship navigation accident investigation report in the step (1) of claim 1 line by line and sentence by sentence, gradually conceptualize and classify the original data through continuous comparison, and remove the initial concept with the frequency below three times, thereby extracting the initial category; the specific content and the method steps of the main shaft code are that the initial category formed by the open code is further refined, different types of the initial category are divided, categories with similar structure, type and logic relationship are classified into one category, and the main category and the corresponding auxiliary category are further distinguished; the specific content and the method steps of the selective coding are that the main category is sorted, the man-made cause (Men), the equipment cause (Machine), the operation cause (Media) and the Management cause (Management) of a 4M theoretical model are compared, the core category is selected by combining the characteristics of ship navigation accidents, the core category is systematically connected with other categories, the relation among the core category and the other categories is verified, and the category which is conceptualized and has not been completely developed is supplemented and completed.
8. The method for analyzing the cause of the ship navigation accident based on the rooting theory and the fact map according to claim 7, wherein the main category refers to safety management, perceptual errors, illegal operation, decision errors, environmental factors, ship equipment factors, personnel factors, improper arrangement of crews, and insufficient navigation plan, and the core category refers to equipment factors, environmental factors, management factors, human factors, and emergency response factors.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210608600.0A CN115033746A (en) | 2022-05-31 | 2022-05-31 | Ship navigation accident cause analysis method based on root-tying theory and incident map |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210608600.0A CN115033746A (en) | 2022-05-31 | 2022-05-31 | Ship navigation accident cause analysis method based on root-tying theory and incident map |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115033746A true CN115033746A (en) | 2022-09-09 |
Family
ID=83123423
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210608600.0A Pending CN115033746A (en) | 2022-05-31 | 2022-05-31 | Ship navigation accident cause analysis method based on root-tying theory and incident map |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115033746A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116756344A (en) * | 2023-08-16 | 2023-09-15 | 中南大学 | Landslide scene body construction method and related equipment for whole process |
| CN117933400A (en) * | 2024-03-21 | 2024-04-26 | 深圳大学 | Knowledge graph-based marine accident analysis method, system, terminal and medium |
-
2022
- 2022-05-31 CN CN202210608600.0A patent/CN115033746A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116756344A (en) * | 2023-08-16 | 2023-09-15 | 中南大学 | Landslide scene body construction method and related equipment for whole process |
| CN116756344B (en) * | 2023-08-16 | 2023-11-14 | 中南大学 | Landslide scene body construction method and related equipment for whole process |
| CN117933400A (en) * | 2024-03-21 | 2024-04-26 | 深圳大学 | Knowledge graph-based marine accident analysis method, system, terminal and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110889556B (en) | Enterprise operation risk characteristic data information extraction method and extraction system | |
| CN110597735B (en) | Software defect prediction method for open-source software defect feature deep learning | |
| CN111428054B (en) | Construction and storage method of knowledge graph in network space security field | |
| CN109657947B (en) | Enterprise industry classification-oriented anomaly detection method | |
| CN115033746A (en) | Ship navigation accident cause analysis method based on root-tying theory and incident map | |
| US20210012210A1 (en) | Techniques for creating, analyzing, and modifying neural networks | |
| CN112579477A (en) | Defect detection method, device and storage medium | |
| CN113254594B (en) | Smart power plant-oriented safety knowledge graph construction method and system | |
| Bukhsh et al. | Maintenance intervention predictions using entity-embedding neural networks | |
| CN113191148B (en) | Rail transit entity identification method based on semi-supervised learning and clustering | |
| CN111882446A (en) | Abnormal account detection method based on graph convolution network | |
| CN104866578A (en) | Hybrid filling method for incomplete data | |
| CN111931505A (en) | Cross-language entity alignment method based on subgraph embedding | |
| CN113779272A (en) | Data processing method, device and equipment based on knowledge graph and storage medium | |
| CN114978613B (en) | Network intrusion detection method based on data enhancement and self-supervision feature enhancement | |
| US20210012209A1 (en) | Techniques for modifying neural network definitions | |
| CN116610816A (en) | Personnel portrait knowledge graph analysis method and system based on graph convolution neural network | |
| CN113254507B (en) | Intelligent construction and inventory method for data asset directory | |
| CN113869052A (en) | AI-based house address matching method, storage medium and equipment | |
| US11615321B2 (en) | Techniques for modifying the operation of neural networks | |
| Chugunkov et al. | Creation of datasets from open sources | |
| CN112035345A (en) | Mixed depth defect prediction method based on code segment analysis | |
| CN111723021B (en) | Defect report automatic allocation method based on knowledge base and representation learning | |
| CN115734274A (en) | Cellular network fault diagnosis method based on deep learning and knowledge graph | |
| CN115204179A (en) | Entity relationship prediction method and device based on power grid public data model |
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 |